Re: Does a Maxtor 5000 LE USB Drive Work under Linux?
This was a thread I started around the first of January and I have good but cautionary news to report. When I first tried the Maxtor drive, it was on a 256-MHZ Dell Pentium mother board with a single IDE controller for the hard drive and CDROM, the usual IBM-style 1.4-meg floppy and a Yamaha sound chip that never quite behaved well under Linux. There was also a 3C509 10Baset Ethernet card and the built-in USB1.1 port. It is a good, but rather ordinary type system by today's standards. The USB driver assigned IRQ9 to service the USB system and maybe that is where things went wrong. I still don't have a good way to know that for sure. All I know is that I got to see failure modes like I had never seen them before. Some disk operations would succeed normally, but most would end up in Never-Never land and one couldn't even tell the system to reboot when that happened because root couldn't even kill the processes. I was in the process anyway of moving all the Linux activity from that system to a newer Dell mother board so I thought I would give the Maxtor another try on the new system. I didn't hold out a lot of hope because the newer system was basically much like the old one except faster and with two IDE controllers and a different Ethernet card. It also turned out the sound system was a better fit under Linux and actually works as it should. It appears that the USB driver chose a better selection of IRQ this time because the Maxtor drive appears to work like a good disk drive should on the new system. The USB port is still a nice slow USB1.1 port so the Maxtor doesn't get to really scream, but file transfers happen at a rate one would expect at a 12-Mb bus speed. It is about like Ethernet over a good LAN using IRQ11 on this system. What this all amounts to is that if you decide to try this type of a set-up, look out for the warning signs that something might not be right. I was lucky for once in that the system I was moving to appears to be working while the one I am retiring (actually returning to its rightful owner) had problems, but it could just as easily be the other way around. In other words, have a plan B. I think Linux represents an incredible human effort and the people who have written these drivers and who do their best to make them automatically do the right thing when faced with an unlimited level of chaos can't be thanked enough. Martin McCormick WB5AGZ Stillwater, OK OSU Center for Computing and Information Services Network Operations Group -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
USB-Based Serial port Modules
I have been meaning to buy a serial port module to connect to the USB port on a Debian system. I should have bought the Edgeport 8-port module when I saw it for sale early this year, because Pricewatch shows only a 4-port model available now or a 16-port unit for $600. Are there any other options? Four ports may not be enough for the various projects I have, but 16 is a bit much at least at $600 when I had planned to spend around $300. I just need to be sure that it will work with the 2.4 kernels. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
USB Serial Driver Isn't quite Working.
I am using the USB mass storage driver on a Maxtor external drive which normally works well. I added a new Edgeport/8 serial converter and plugged it in to the second slot of the USB port. The documentation in /usr/src/linux/Documentation/usb-serial.txt says that the generic serial driver supports the Edgeport/8 product. When I plug it in, the following lines get added to syslog: hub.c: USB new device connect on bus1/1, assigned device number 3 usb.c: USB device 3 (vend/prod 0x1608/0xf) is not claimed by any active driver. Do I need to add any other modules to the kernel when compiling? Disconnecting the Edgeport/8 produces: usb.c: USB disconnect on device 3 The mass storage driver continues to work with the serial port unit in or out. At least the new generic serial driver doesn't appear to have done any harm. Martin McCormick WB5AGZ Stillwater, OK OSU Information Technology Division Network Operations Group -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: SCSI emulation for IDE CD-writer
Steve Doerr wrote >I overlooked the SCSI cd support when I built my kernel. > >It's working perfectly now. I did almost the same thing on a system at work except I forgot the Generic SCSI support. I had a system that looked like it should work if you did the cat /proc/scsi/scsi tests, and everything else looked good, but you sure couldn't write a CD. It complained that the device was missing /dev/sg0 cdrecord -scanbus told me to run cdrecord -scanbus, etc. I really felt stupid after discovering what I had done. After I built a new kernel with the Generic support, it burns CD's like a fire bug. Martin McCormick WB5AGZ Stillwater, OK OSU Center for Computing and Information Services Network Operations Group -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
What Happens When the USB Mass Storage Driver Doesn't Work?
I am trying out a new external Maxtor 5000 LE hard drive I bought with the understanding that I can return it if it won't work under Linux. I built a 2.4.19 kernel, adding support for the usb port, the mass-storage driver and SCSI disk support. I had SCSI emulation working successfully to enable use of a CDRW drive. What I have succeeded in doing is killing the SCSI emulation for the IDE CDRW drive. All the logs report that the USB hub driver and mass-storage USB driver are successfully registered. kernel: usb.c: registered new driver hub kernel: Initializing USB Mass Storage driver... kernel: usb.c: registered new driver usb-storage kernel: USB Mass Storage support registered. Now, no SCSI devices show up in output from cat /proc/scsi/scsi I used to see the emulated SCSI channel for /dev/sr0, but that is gone as long as I have SCSI disk support. I would think that if the new drive was the problem, I should see some kind of SCSI errors reported or experience long pauses and or time-outs, but everything is strangely silent after the initial syslog messages upon boot-up. I would expect that I should get some kind of access attempt messages even if the drive was not connected to the port. I did enable verbose error reporting for the SCSI system, but those messages I showed here are the only evidence that anything happened with the USB ports at all. So, my questions are: Has anyone gotten one of these nifty-looking drives to talk to a Linux system? What kind of messages should I see, if any, when the usb drivers are working right but the drive isn't? This drive says it works with Mac OSX and, of course, Windows. I need to know if it should work under Linux fairly soon because of the return question. For background, the kernel had SCSI emulation enabled for the CDRW drive, Generic SCSI support and SCSI CDROM support. I also have to have an append statement in lilo.conf to map hdc to SCSI. For the USB hard drive, I added SCSI disk support, usb support and the usb mass-storage driver. That, as previously stated, is when I broke SCSI emulation. There shouldn't be any conflict with the naming of SCSI devices since it looks like the disk should show up as /dev/sda if it ever does.:-) Many thanks for any constructive suggestions and a happy New Year to all. Martin McCormick WB5AGZ Stillwater, OK OSU Center for Computing and Information Services Network Operations Group -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: IDE SCSI Emulation and USB problem
I have been doing some further tests about my usb problems and I found out that my testing procedure is flawed. Every new kernel I built was built using bzdisk as the last step in the process so I left the kernel on my boot drive alone and was using a new kernel built on a floppy disk. I believe that that kernel uses a different lilo.conf file because that was part of the reason why I see no SCSI emulation for the IDE CDROM. I bit the bullet and built a new kernel with the IDE SCSI emulation as before and the usb port mass storage driver support. This time, I still have the CDROM as /dev/sr0 or /dev/scd0 so that is good. What I still don't seem to have is anything happening with the usb port and mass-storage driver. The syslog shows that the usb drivers installed, registered and initialized, but there are no new devices. I am testing a new Maxtor 5000 LE external drive with this system so I have the two unknowns of the new usb installation plus the new drive which may or may not work with Linux. The drive works with Windows and Mac OSX so it is not known whether or not it will work here. Shouldn't there be some kind of dialog regarding the usb activity when the system boots besides the simple fact that the usb support is there? I even temporarily took out the append command in lilo.conf that maps /dev/hdc to scsi and the SCSI did go away. The usb support, however, does not appear to have added any new devices with the only thing at all being a usb directory in /proc/bus which is empty. Does it matter which usb port I connect to on the mother board? Many thanks. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: IDE SCSI Emulation and USB problem
According to the documentation for USB, the drives that should show up on your system are /dev/sda or /dev/sdx if you already have a /dev/sda or /dev/sdb, etc. In other words, the USB devs should start at one above any pre existing SCSI disks. This means that your /dev/scd0 should not conflict with any of the USB-based SCSI emulations. "Charlie Reiman" writes: > > >> -Original Message- >> From: Aryan Ameri [mailto:[EMAIL PROTECTED]] >> Sent: Tuesday, December 31, 2002 10:25 AM >> To: debian users >> Subject: IDE SCSI Emulation and USB problem >> >> >> Hi there: >> >> In order to be able to burn CDs using my ATAPI DVD/CDRW drive, I did the >> followings to get ide-scsi emulation working. >> >> I first loaded the ide-scsi module using modconf. >> Then Added the following lines to /etc/modutils/actions >> >> "alias scd0 srmod >> alias scsi_hostadapter ide-scsi >> options ide-cd ignore=hdc" >> >> Then added the following line to /etc/lilo.conf right after the >> "root=" line >> >> "append="hdc=ide-scsi" >> >> I rerun lilo, then rebooted and changed /etc/fstab to mount >> /dev/scd0 instead >> of /dev/cdrom. >> >> I thouht that was it, I started burning CDs, and everything was >> OK, untill I >> saw that I have two problems >> >> 1. My USB programs can't detect USB devices, e.g gtkam can't >> detect my digital >> camera, and my mp3 player program can't detect my mp3 player. > >Most USB devices that act as filesystems use a SCSI emulation layer as well. >I believe it also defaults to using /dev/scd0. You might want to try >changing the IDE device to a different /dev/scd file. > > >-- >To UNSUBSCRIBE, email to [EMAIL PROTECTED] >with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED] > -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Does a Maxtor 5000 LE USB Drive Work under Linux?
I presently have 2 unknowns I am trying to resolve. I have installed USB support and the USB mass-storage driver on a Debian system using the 2.4.19 kernel. The USB stuff installs with no complaint at all. I power up the Maxtor and boot the system but no new drives show up in the empty usb file system that can now be found in /proc/bus. I also don't see any kind of errors in any log files so I don't know if the Maxtor can't talk to Linux or if the usb drivers aren't working. The drive is new and I can return it to the store where I got it if it isn't going to work, but I need to do that soon or I must eat it. Does the USB system complain if it can't talk to any drives? Things are just too quiet with the only log entries being a successful report that the USB module installed as well as the mass-storage driver. My hunch is that the USB is okay but it won't talk to this drive. I just want to be sure I have tried everything before I take it back. I could sure use the storage. If anyone has seem my earlier post under a different subject, the loss of SCSI emulation of IDE drives appears to have been a false alarm. I now have both USB and SCSI emulation here, just no new blank drives called /dev/sda. -- Martin McCormick WB5AGZ Stillwater, OK OSU Center for Computing and Information Services Network Operations Group -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: IDE SCSI Emulation and USB problem
Aryan Ameri writes: >So where is the conflict from? I am not sure. I thought I had the same problem you reported when I was (still am) trying to install a USB-based external drive on my system. It turned out that my practice of building test kernels on floppy disk and running them by booting from the floppy caused a different lilo.conf to be used that didn't have the append directive in it to map the IDE CDROM drive to the SCSI emulation. As I understand it, all the SCSI CDROM drives should show up in your list of SCSI devices with names that are like /dev/sr0 or /dev/srx depending upon how many drives you have. It doesn't matter if they are true hardware-based SCSI devices or the SCSI emulation is grabbing them. CDROM's and hard disks each have their own nomenclature and they should co exist peacefully.:-) If I am wrong, somebody please tell me so for I hate to lead others astray. One way you might get this problem is if your CDROM is somehow being mistaken by the system as a writable hard disk. I did see a warning on a Linux installation program that indicated this could happen on some systems and that might be why they are showing up wrong although one would think that you would just see more hard disks than you should have and less or no CDROM's. Martin McCormick WB5AGZ Stillwater, OK OSU Center for Computing and Information Services Network Operations Group -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Does a Maxtor 5000 LE USB Drive Work under Linux?
I want to thank all those who have offered suggestions. I have a classic good news/bad news progress report. The good news is that the drive will work under Linux. It appears to come with a Windows file system, but fdisk and mkfs took care of that. My problem was that I ended up having no interface driver. I had mis understood some of the documentation that describe use of UHCI and OHCI drivers. I thought that the USB support was all you usually needed along with the mass-storage drivers. It turns out that on my machine, you need the alternate driver called UHCI. This produces huge output in the syslog when it finds the drive and it even reports that it is a Maxtor 500 LE. I was extatic and really whooped it up when that happened. Then the bad news started happening. The drive works, but I have something seriously wrong somewhere. Firstly, I have tried both the UHCI and OHci drivers to see which one is the best fit. The OHCI driver didn't produce so much as a single syslog message and there was no connectivity. The UHCI driver finds the hard drive, but then things get strange. I seem to be able to mkfs pretty well if I use /dev/sda1. I cam copy small files in to and out of it perfectly. If I copy a big file such as a 50-meg audio file which is why I installed it in the first place, I can sometimes get about 32 megs copied before 1 of 2 things happen. The fsck or large file copy falls silent. Opening another login session and running a ps shows the original process plus a child spawned by fsck which usually shows up as having a code of d that sometimes changes to r and then back to d. As time passes, it is much more d than r and you can't kill it with an atomic bomb. Kill -9 does nothing to it and if you finally give up and are lucky enough to have an open session to root, reboot starts and then hangs because it can't even kill that process. The only thing that will kill it is literally turning off the power to the system or, of course, hitting the Reset button. Then, you must do fsck on root to get back to a working system. I don't see any syslog errors, but I just know this is some kind of resource depletion problem. The drive is working but my computer which has been rock solid up to now is almost guaranteed to need a shut-down or hardware reset to bring it back to life after a large file copy or fsck. I am actually about to move everything to a faster system with twice the memory of the present system, so I don't know if the problem will follow, but it would be nice to know if anybody has any theories as to what I might look at. I have 64 megabytes of RAM and a 64-meg swap partition and the main hard disk ranges from 60 to 80% full so it isn't running out of room for temp files as far as I can tell. The other one of the two bad things that happens is that you suddenly can'tspawn any new logins or one that is open to root freezes up and absolutely nothing continues to work. A look at processes when the system is sick but not dead yet shows the parent and child I mentioned plus the usual assortment of Linux processes, but nothing particularly odd. Thanks for any ideas as to what is happening and how to fix it. I am so close to success, but still not there yet. Martin McCormick WB5AGZ Stillwater, OK OSU Center for Computing and Information Services Network Operations Group Nicos Gollan writes: >-BEGIN PGP SIGNED MESSAGE- >Hash: SHA1 >You should perhaps consider writing a mail to Maxtor support if you >haven't done so already. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Does a Maxtor 5000 LE USB Drive Work under Linux?
I am the person who started this thread originally and I have spent a bit over three days thinking I was about home free on getting this drive to work. On the good advice of another list member, I contacted the Maxtor Web site with the idea of discussing how to make it work, but I didn't need to because somebody has already asked. Here is the answer lifted from their web site. Technical Bulletin Answer ID 1271 Product Personal Storage 5000DV 5000LE 5000XT Category Compatibility Date Updated 10/17/2002 02:36 PM Printer Friendly Version of This Answer Print Answer E-mail This Answer E-mail Answer Will the Personal Storage 5000 work with Linux? Description Personal Storage 5000 LE/DV/XT Functionality in a Linux Environment. Answer Hardware drivers normally come from different operating system manufacturers. The Linux development community has not yet, as far as we are aware, developed drivers to allow Linux to be used with our 1394 and USB interface. I guess the answer is, "not yet." This is a lot better than no. If I knew just a little more than I presently do, I would volunteer to tackle this problem, myself, but I don't yet really know how the USB or 1394 interfaces work. This has all the earmarks of either a hand-shaking protocol gone awry or an interrupt handler that may not be returning cleanly from an ISR since it is so random and it also seems to be related to the size of the data transfer. One writes or reads a small file and it works. Big operations are sure to fail with the only question being whether just the process hangs or the whole system dies. If it is a hand-shake problem, something is blocking and waiting for something that never happens or happens so quickly that it gets missed. If it is an ISR problem, maybe data are being lost when returning or multiple interrupts are happening and blowing out the stack. I have written some assembler programs before and I am aware how hard it can be at times to debug events whose timing can be effected by tracing program flow to the point that it is very hard to replicate the same conditions that are causing things to blow. If there is a better place to post this information, please tell me. If given enough time, I might figure it out by the end of the decade, but I hope there are folks who are already on the ground running with it. Martin McCormick WB5AGZ Stillwater, OK OSU Center for Computing and Information Services Network Operations Group -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Minimal Woody Installation Disk and Serial Console
I am using the Serial console method of installing Woody on a new, or at least, new to me, Dell system. For those who haven't done this, you just insert the Debian boot CD and boot the system but type the following where most people just hit Enter: linux console=ttyS0,9600n8 and the Com1 port begins to send and receive all of the installation messages. This works for me until I get to the end of the first part of the configuration sequence at the point where you have to configure the Base system. At that point, you have to reboot the system and that's where things go wrong. Interestingly enough, if you build a system like that, the inittab file gets modified to make ttyS0 or which ever port you used, in to a tty login port which is very neat. Now for the bad news! What I immediately get is the remainder of the configuration sequence where one sets the time zone, host name, network acquisition method such as dhcp or ppp, etc. The problem is that now, the system seems to have forgotten that I am using a VT100. I get a very bad screen full of escape sequences and garbage like one gets when looking at a vt100 session with the wrong terminal type. When I get to the screen to set the root password, I am in an endless loop in which the system tells me that I entered an empty password which is bogus. It then proceeds to be stuck in that mode forever until I die and stop trying to enter a password.:-( GRRR! I don't fully understand what is wrong since I use Kermit's vt100 emulation on the P.C. which serves as the terminal, but I am awash in 8-bit characters and their ill effects on Kermit's operation. I can actually continue to muddle through the mess to set the time zone and system name, but that password screen is as good as a lockup since you can't get past it. I have installed earlier versions of Linux on other systems and run in to some of the same problems, but never this bad. Any ideas as to how to make the base configuration send proper vt100 codes and let me finish this installation? For computer users who happen to be blind, this is one of the best methods for installation since everything you need is on the CD, but I am stuck. The last thing that got successfully set was the time zone and I was able to bring up the network card, but all that gets lost each time I have to bring it down since it is dhcp and there isn't enough of a system up yet to safely put it on the network anyway. I did make a boot floppy when given the opportunity and it works perfectly and puts me right back at the start of the remainder of the configuration. That is fine if I can ever get proper communication with it and complete it. Any ideas please? Martin McCormick WB5AGZ Stillwater, OK OSU Center for Computing and Information Services Network Operations Group -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Minimal Woody Installation Disk and Serial Console
I have partly solved my own problem. The bash-type shell which is part of the installation software is just what I need to get a shot at modifying the inittab file. It turns out that a special inittab file is installed just for the installation process and then the real inittab is put in after one finishes all the configuration. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Snort Messages may not be Telling Me Much.
I've got snort installed and running on Debian3.0. It runs fine but I never get any thing in the report Emails that I receive each day. The messages have a suspiciously empty look to them like a form that should be filled in but isn't. I have run snort in the past in a fully-capturing mode and there are all kinds of the usual monkey business going on out there so I know that in 3 months, I should have seen something. After 92 days, every single message looks like the following example: (Message inbox:152) From:root <[EMAIL PROTECTED]> Subject: [SNORT] systemname daily report The log begins from: :: The log ends at: :: Total events: 0 Signatures recorded: 0 Source IP recorded: 0 Destination IP recorded: 0 The number of attacks from same host to same destination using same method = # of attacks from tomethod = Percentage and number of attacks from a host to a destination # of %attacks from to Percentage and number of attacks from one host to any with same method == # of %attacks from method == Percentage and number of attacks to one certain host = # of %attacks tomethod = The distribution of attack methods === # of %attacks method === End of example --- I installed snort by using dselect and it is looking for all the standard attack signatures that come with the distribution. I certainly don't want trouble, but I think I am missing activity that is going on. The network it is sniffing is the correct network for my installation and, in the past, I saw traffic when using that definition so I am not sure what is going on. Any constructive ideas are welcome. By the way, this particular message is not coming from the system in question. Martin McCormick WB5AGZ Stillwater, OK OSU Information Technology Division Network Operations Group -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Snort Messages may not be Telling Me Much.
David Mandelberg writes: >I'm having the same problem. From what I hear, a solution has something >to do with acidlab, but I'm not sure. Thanks. I have another Linux system at work running the same version and I noticed it also has the same problem. Martin McCormick -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
gunzip and Unexpected EOF
before rebuilding a system after an upgrade that went terribly wrong, I did the following in order to save kernel sources and configuration files: tar cf - . |gzip >../otherfilesystem/root.tar.gz This worked fine on the home file system and I recovered all the /home files but when I went to do tar zxf otherfilesystem/root.tar.gz, things seemed to start for a while and then disaster!!! gzip: stdin: unexpected end of file tar: Read 4489 bytes from /mnt/root.tar.gz Apparently a special file or something in / makes gzip stop before decompressing most of the 900 MB file. Is there any way to cause it to go any farther? I am not trying to restore every byte of the old file system, only selected parts such as /usr/local/src, /usr/src and various crontab files. Of course, the creation of the archive left no hint that it was not going to unpack later. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: gunzip and Unexpected EOF
Almut Behrens writes: >I'm afraid your tarball got corrupted... In that case - I'm sorry to >say - you're almost certainly outta luck. After creating a tar.gz >file, the _whole_ tarfile has been piped through gzip, so if any part >of the file gets corrupted afterwards (however tiny - a single bit is >sufficient), the gzip algorithm will not be able to recover any data >later in the stream... > >My personal advice to avoid such desaster in future: use 'afio' >instead of tar + gzip. afio compresses each file seperately, so in >case of partial corruption if the archive, most of the remaining files >can still be recovered... Thank you. I was kind of afraid of that. Fortunately, I had backups of the /usr file system and /etc which were from mid May and haven't changed too terribly much so I think I am not as hosed as I could be. Thanks again for the advice. Martin McCormick WB5AGZ Stillwater, OK OSU Information Technology Division Network Operations Group -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Fixing Mistake in grub config.lst; Need Rescue Disk
I successfully installed Debian from a CDROM burned from the image debian-31r0a-i386-binary-1.iso I then tried to install a 2.6.5 kernel on that system and must have incorrectly modified /boot/grub/menu.lst because the system doesn't boot. I did think to make a backup of the original menu.lst file but then I discovered that if I boot from the Debian installation CDROM, I can't seem to use the shell one can spawn to mount any file systems. Is it possible to use that disk as a repair disk? If so, what am I missing? The drive is already formatted and only needs to be mounted so I can do the 5-second mv operation that I now have spent about 6 hours trying to accomplish. If that disk isn't a good repair disk, is there an image I can download which understands the ext3 file system? One other thing that may complicate this request is that I need to do this via the serial console. The installation CD I burned from the image indicated above lets one do this all right so what I need is something like that that will let me directly mount the root file system. Thanks for any help. Martin McCormick WB5AGZ Stillwater, OK OSU Information Technology Division Network Operations Group -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
SCSI Emulation Directives in grub/menu.lst
I am rebuilding a 2.6.5 Debian system and everything is
working properly so far. The system had ext2 file systems and used
lilo and now has a ext3 journalling file system on the master drive.
It also uses the grub boot loader and appears to be healthy.
Before I rebuilt it, I used scsi emulation to make CD burning work and
have been lucky in that both a Sony and Plextor IDE drives have
not missed a beat when burning CD's. The lines in /etc/lilo.conf are:
append = "hdc=scsi"
append = "hdd=scsi"
A message that SCSI emulation is deprecated for CD burning
pops up, but I have also heard of problems if one uses /dev/cdrom so I
would like to continue to use scsi emulation as long as I use a 2.6
kernel.
Is there an equivalent grub configuration line for menu.lst
that does the same thing?
If I can't use scsi emulation any longer, will cdrecord and
cdparanoia actually work since they appear to rely on scsi protocols?
I did read the grub manual, but maybe missed the obvious.
Nothing jumped out at me.
When building the new system from the latest DEbian
distribution, I had a choice of grub or lilo but grub was the first
boot loader presented. Lilo fails to install so I went with grub
since that appears to be the general direction of progress.
Thanks for your ideas.
Martin McCormick WB5AGZ Stillwater, OK
OSU Information Technology Division Network Operations Group
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Will cdrdao actually work on a 2.6.5 Kernel?
The system in question has 4 IDE interfaces and the CD drives map to /dev/hdc and /dev/hdd. After trying several syntax possibilities like: cdrdao read-toc device:ATAPI,#,#,# toc_testfile and, after reading more documentation: cdrdao read-toc device:ATA,#,#,# toc_testfile I had a glimmer of hope. When I gave the device spec of ATA:0,0,0 the output from cdrdao said it couldn't open /dev/hda which made me realize I had the wrong unit specification and that I should probably try ATA:2,0,0, etc. That also produced an error that cdrdao couldn't open /dev/hda which tells me something is not right. A search of 'HOWTO' documents for burning CD's says to try using cdrecord to learn drive options. I installed the modified cdrecord from the Debian packages and got nada. It, in its own charming way, said "I don't know what you're even talking about." Actually, it wanted to see SCSI devices and couldn't find any except for a USB-based external hard drive which is connected to the system, not the IDE CD drives I was expecting. The syntax for cdrdao may be wrong or it may just not work on this system, but it is certainly not unusually configured. The first IDE controller has two hard drives which show up as hda# and hdb# depending upon what partition one means. The only good news so far is that cdparanoia from the debian packages has no trouble at all with /dev/hdc so I'm almost back to where I was in the rebuild and modernize process. Thanks for any ideas. Martin McCormick WB5AGZ Stillwater, OK OSU Information Technology Division Network Operations Group -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
BitTorrent Clients and X
Does BitTorrent require X to run under Linux? I was going to download a file that appears to exist only in .torrent form and was shunted to the BitTorrent web site to download a .deb package. I figured this was a good time to learn about BitTorrent so I downloaded the package and nothing much else really happened. This system does not run X and I think what really happened was that the package unpacked and couldn't install. A look at the BitTorrent web site mentions the GUI in system requirements, but I am not sure if that is optional or what. In my case, a required GUI equals no application period. Martin McCormick WB5AGZ Stillwater, OK OSU Information Technology Division Network Operations Group -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: BitTorrent Clients and X
Sven Hoexter writes: >Look out for btdownloadcurses.py and the other btdownload*.py programs >installed in /usr/bin Thank you and also thanks to the person who recommended apt-get install bittorrent. I actually had tried that but what was truly wrong was that I hadn't updated the package lists recently. Bittorrent showed up after I did that and I needed to fix a bunch of dependencies which is what killed the attempt to install from the BitTorrent web site. Thanks for your help. Martin McCormick WB5AGZ Stillwater, OK OSU Information Technology Division Network Operations Group -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
dselect and a Recent Experience
In the spirit of helpfulness on this list, I want to repay some of the assistance I have gotten from others, here. This message is a warning, not a gripe. I think Linux software is astoundingly robust. Like anything, there are gotchas. I am not sure exactly what happened, but I last week trashed my system at home while doing a rather innocent upgrade based upon what dselect computed needed to be done while installing a single package. I wasn't paying attention closely enough because things usually work perfectly and what happened was that a large number of packages got removed and the system ended up not even being bootable. Fortunately, I do /home backups on a daily basis and I was even able to use a rescue disk to tar /home so I didn't loose data. The new version of my system now has ext3 file systems for everything and grub as the boot loader. the only real loss was time. Today, here at work, I was going to install bittorrent on ta Debian system and I used dselect to list the packages in order to find bittorrent. It found it and I started to install from there except I saw that dselect was also going to whack about the same number of packages here that it did when I was on my system at home. Both are very stable woody-installed systems. I was actually awake at the switch this time so I aborted the tragedy before it happened. I used apt-get install bittorrent since I knew I had it in the package lists and it appears to have worked fine and corrected the dependencies needed to install bittorrent. dselect has, for the most part worked very well so I think there is a possible bug somewhere to cause it to get confused as to what it is supposed to delete, but everyone needs to pay closer than usual attention to what dselect is doing if or when you use it. Martin McCormick WB5AGZ Stillwater, OK OSU Information Technology Division Network Operations Group -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: dselect and a Recent Experience
Oliver Elphick writes: >Does your /etc/apt/sources.list reference woody or stable? > >They used to be the same, but now there is a new stable, sarge. So if >you reference stable, a lot of your packages are now regarded as >obsolete. You should either reference woody explicitly or do a complete >upgrade (recommended!) using the instructions in the release notes at >http://www.debian.org/releases/stable/i386/release-notes/index.en.html Excellent question that hadn't even occurred to me! Here's the important lines in sources.list deb http://http.us.debian.org/debian stable main contrib non-free deb http://non-us.debian.org/debian-non-US stable/non-US main contrib non-free deb http://security.debian.org stable/updates main contrib non-free deb http://http.us.debian.org/debian/ stable main deb http://non-us.debian.org/debian-non-US stable/non-US main That's probably the problem right there. I also had deb http://www.carezia.eng.br/debian woody main This was for a modified version of snort which worked well on my system at the office, but never quite worked right on the home system so I took that source out of the sources.list at home about 6 or 7 months ago. Both systems swarmed on me in the same way so I think you hit the nail on the head. The system at home is probably now safely upgraded since I whacked the whole OS and started from scratch with the newest Debian netinstall CD. I should make sure the work system gets upgraded soon and that will probably make dselect safe to use again. Many thanks. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Movie players
Piero Piutti writes: >All you need is mplayer, the w32codecs and mozilla-mplayer (aka >mplayerplug-in). It will work as a charm: tested with the Quicktime trailers >on Apple's website. I second that. I installed MPlayer-1.0pre7 and the codecs directory that one should also install and am amazed how well it works. I mainly use it to listen to streaming audio from various radio stations. As a computer user who happens to be blind, I do not run X yet because it is not quite ready for prime time so I was amazed to find that I could even listen to some TV audio. I tried solent.tv which is in the Isle of White in the UK and was able to listen to the audio of a recent newscast. It was a .wvx file and I have no idea whether or not there are PAL/NTSC/SECAM issues or not. The sound was not terribly good but it didn't sound broken in any way. The codec gave it a hollow well sort of effect, but the voices and music were the correct pitch and probably sounded like that on any computer accessing the broadcast. MP3 streams and Windows audio streams sound anywhere from perfect to pretty good depending upon the sampling rate. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
nmh replcomps configuration for Replies to Mailing Lists
I have been attempting to adjust my replcomps file so that
replies to mailing list mail are more likely to go to the list rather
than the individual poster. Is there any way to cause repl to respond
correctly to a line like:
list-post: <mailto:[EMAIL PROTECTED]>
I think the mailto directive is what's confusing things.
If I put it in replcomps, repl finds it but can't understand its meaning.
Of course, several lists I am on appear to use that header as the only
reliable header that could control the construction of the reply address.
Thanks for any useful ideas.
Right now, the pertinent line in replcomps looks like:
%(lit)%(formataddr %<{delivered-to}%?{X-Original-To}%?
{mail-followup-to}%?{sender}%?{return-path}%?{mail-reply-to}%?{reply-to}%?{to}%?{from}%>)\
Line was broken for formatting purposes.
Martin McCormick WB5AGZ Stillwater, OK
OSU Information Technology Division Network Operations Group
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Is there a Way to Read PDP-11 Floppies?
I found some documentation about how to use dd to analyse an unknown disk. In this case, the disks are boot loaders for some 3Comm serial communications servers. The procedure I read about tells one to do something like dd if=/dev/fd0 of=/tmp/testfile count=1 # If it works: getfdprm# This will report what geometry the disk has file /tmp/testfile # This may indicate the type of file system mdir a: # Check for an MS-DOS file system tar tvf /dev/fd0# Check for a tar archive cpio -itv < /dev/fd0# Check for a cpio archive e2fsck /dev/fd0 # Check for an "ext2" file system I did all that and found that the disk is DS HD sect=18 /tmp/testfile: PDP-11 UNIX/RT ldp I can use dd to seemingly copy the entire disk image with no trouble or at least not one error. dd bs=2x80x18b of=/dev/fd0 if=/tmp/floppy.image If I try to write floppy.image back to /dev/fd0 with a blank but formatted disk in it, the resulting disk appears to have the same data on it but it won't boot and the Linux mount command complains about bogus FAT's. Is there any sort of Linux analysis tool that works in the command mode that might tell me what is wrong with the copies I am making? These disks are not copy-protected at all and one can use the comm server to make new copies but this is not as convenient as using a computer to manage backups. It would be nice to be able to mount the disks, but being able to make a binary copy to and from an image would be fine, also. Also, the problem may lie in the way I use superformat to format a totally blank diskette. I do: superformat /dev/fd0 --verify_later sect=18 cyl=80 I think what I need is to format all the tracks with no FAT. The PDP-11 disk has a different FAT structure, at least that is what I think is happening. If anybody has any suggestions as to how I can possibly complete the copy process, I am extremely thankful. Martin McCormick WB5AGZ Stillwater, OK OSU Information Technology Division Network Operations Group -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Another shell scripting question
A response to another poster peaked my curiosity. > Incidentally, there is no reason to make it a bash > script rather than vanilla sh, and you can simplify the script by using > exec: I have been writing shell scripts for a bit over fourteen years so I am not new to this, but I unfortunately get caught up in solving the problem of the moment and therefore fail to learn as much as could be learned. Is it just more efficient in resources to use plain #! /bin/sh rather than bash? I usually always write my shell scripts with the Bourne shell unless I discover that something doesn't work in sh but does work in bash and then I use bash. Some arithmetic functions seem to be easier to code in bash but that might also be because I didn't know the magic syntax to make them work in sh. Martin McCormick WB5AGZ Stillwater, OK OSU Information Technology Division Network Operations Group -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Swiss Army Knife of Sound has me Baffled.
I have a little shell script called fullstereo which works fine. It's short so I'll show it to you. It records sound from a Creative Labs usb sound card which is probably much happier on a Windows box but that's not where I need it. It has only 1 sampling rate that works under debian and that is 48 KHZ. I want 44.1 KHZ and I get 2 good stereo channels with the following script: #! /bin/sh cd ~/tmp filename=$1.wav echo $filename arecord -D hw:1,0 -r 48000 -d $2 -c 2 -f S16_LE - \ | sox -t wav - -r44100 $filename $2 is a value in seconds, minutes or hours and it goes like a house on fire with no problem. So, I want a mono version. Let's try this: #! /bin/sh cd ~/tmp filename=$1.wav echo $filename arecord -D hw:1,0 -r 48000 -d $2 -c 1 -f S16_LE - \ | sox -t wav - -r44100 $filename This hasn't worked yet. It is identical to the good version but for c 1 as the number of channels to arecord instead of c 2 . When running that script, this is what happens: $ ./highmono testsound 30 testsound.wav Recording WAVE '-' : Signed 16 bit Little Endian, Rate 48000 Hz, Mono #Good so far and then arecord: set_params:1345: Channels count non available sox FAIL formats: can't open input `-': WAVE: RIFF header not found Darn! A couple of microseconds before that, the same process understood what I asked it to do. When the -c parameter was set to 2, it understood what to do with the standard output but with c set to 1, it did what you see above. I can't seem to figure out why 1 channel output is any different from 2 channel except that the wav header should be pretty much the same with a channel count of 1 instead of 2. The man page for arecord even says the following: -c, --channels=# The number of channels. The default is one channel. Valid val‐ ues are 1 through 32. Theoretically, one could leave out -c 1 and it should go ahead as if that flag was actually there but it breaks instead. This might also be because one can not set any parameters without the Windows driver and I guess I could leave it in stereo mode and tell sox to mix the left and right channels which is fine with me but kind of clunky. I am asking the list's wisdom in case there is a better way to produce the mono channel. Thanks. Martin WB5AGZ
Re: Swiss Army Knife of Sound has me Baffled.
Dan Ritter writes: > > arecord: set_params:1345: Channels count non available > > sox FAIL formats: can't open input `-': WAVE: RIFF header not found > > Keep recording 2 channels. Your hardware doesn't like it > otherwise. > > You can ask sox: > > sox -t wav -c1 - -r44100 $filename > > to merge the channels to a single output channel > > I believe, but have not tested, that a final argument to sox of > > remix 1 > > will get you just the left channel, or remix 2 for the right > channel. > > -dsr- I haven't tried anything new yet but that sure looks right and will probably work. Thanks to all who responded. This is what I like about this list as it presently exists. I had figured out that the problem was with the hardware which is a fairly decent device but severely limited in Linux because a lot of it's power comes from a proprietary driver like a number of other devices such as winmodems, etc. Then I read this: Nicolas George writes: > If you read carefully, you will notice that your problem is not with > sox, it is with arecord: sox only fails because arecord fails and does > not produce anything. > > arecord itself fails because "channels count not available": the sound > device you are using does not like mono, it obviously like only stereo. > > You have two options: > > - Use a sound device that supports mono. Since you cannot change your > hardware, it needs to be done in software with an ALSA plugin. This is > all streamlined, you only need to write plughw instead of hw. > > - Keep recording in stereo, and have sox convert it to mono before > saving. You will need to peruse the manual page of sox to know the > option, I only ever use ffmpeg for obvious reasons. > > Note that: > > - If you use plughw instead of hw, then the ALSA plugin can also > resample, and therefore you no longer needs sox at all. > > - sox is capable of recording from ALSA directly, with something like > "-t alsa plughw:0". > > Regards, I actually did get a message from arecord about using a plugin and you have answered the question as to what needs to be done to make that happen. I had not even thought of the plugin solution so this is a whole new possible route to success which I didn't even think of before. David Wright also writes: > Try -r 8000, or leave out -r and -c entirely, and you should > succeed. But I doubt that's really your priority, is it? > > > […] and I > > guess I could leave it in stereo mode and tell sox to mix the > > left and right channels which is fine with me but kind of clunky. > > That's why it's the "Swiss Army knife". After mixing the two channels, > you can write to a single-channel WAV file. > > What you can't do is tell /arecord/ to use an ADC that doesn't exist. > > Cheers, > David. I am in no great hurry so I will probably try both the plugin solution and the sox coding and save them both for later as the idea is to end up with something that is both efficient and useful. Thanks, everybody. Martin McCormick WB5AGZ
Proprietary USB Drivers; Ya Gotta' Love'em.
This is an embedded usb serial port in a radio receiver which works in MS Windows10 and I was hoping to write some control routines for a Raspberry Pi running buster and it does the following: Sep 22 20:17:15 rpi2 kernel: [551843.541477] usb 1-1.2: new full-speed USB device number 6 using dwc_otg Sep 22 20:17:16 rpi2 kernel: [551843.674681] usb 1-1.2: not running at top speed; connect to a high speed hub Sep 22 20:17:16 rpi2 kernel: [551843.684432] usb 1-1.2: New USB device found, idVendor=0c26, idProduct=002b Sep 22 20:17:16 rpi2 kernel: [551843.68] usb 1-1.2: New USB device strings: Mfr=1, Product=2, SerialNumber=3 Sep 22 20:17:16 rpi2 kernel: [551843.684452] usb 1-1.2: Product: IC-R30 Sep 22 20:17:16 rpi2 kernel: [551843.684460] usb 1-1.2: Manufacturer: Icom Inc. Sep 22 20:17:16 rpi2 kernel: [551843.684468] usb 1-1.2: SerialNumber: IC-R30 16002557 Sep 22 20:17:16 rpi2 kernel: [551843.727911] cdc_acm 1-1.2:1.0: Zero length descriptor references Sep 22 20:17:16 rpi2 kernel: [551843.727940] cdc_acm: probe of 1-1.2:1.0 failed with error -22 Sep 22 20:17:16 rpi2 kernel: [551843.728056] usbcore: registered new interface driver cdc_acm Sep 22 20:17:16 rpi2 kernel: [551843.728061] cdc_acm: USB Abstract Control Model driver for USB modems and ISDN adapters Sep 22 20:21:11 rpi2 kernel: [552078.765119] usb 1-1.2: USB disconnect, device number 6 So much for that. There is one other possibility. It is possible that one can go to one of the menu options and flip a virtual switch that will take the serial data that should have been on /dev/ttyACMx and put them on a logic-level serial line that is accessible through the right stereo channel of a 3-conductor jack. Whether or not that works is beyond the scope of this list so my question is whether or not there is a wrapper program of some kind in Linux that can fool an application like the serial port module that tried to run here in to thinking that it is home and among friends when it's really been kidnapped. The possible use of a logic-level serial line is probably what I will need to do but that means that if one resets the radio to factory defaults, one will have to go in to the menus on the radio's screen and restore the virtual switch setting. As a computer user who happens to be blind, that's a major pain as well as in a case where the radio is remotely located and one is operating it headlessly. I smiled a bit when reading the syslog admonition to connect to a high-speed hub. That would be quite a trick. Thanks for any constructive suggestions. Martin McCormick WB5AGZ
Re: Proprietary USB Drivers; Ya Gotta' Love'em.
Charles Curley writes: > Possibly a known kernel bug. > > https://forums.opensuse.org/showthread.php/538695-USB-driver-Zero-Length-Descriptor > > https://lore.kernel.org/lkml/1507709452-31260-1-git-send-email-msa...@iotecha.com/ > > -- > Does anybody read signatures any more? > > https://charlescurley.com > https://charlescurley.com/blog/ Thank you. That's got to be it. I've been doing debian Linux since right around 2000 but I must admit that I don't know the easiest way to include the patch referenced in the second link. I will need to do that on an i86 box plus the Raspberry Pi I first discovered the issue on since I am apt to have the radio connected to either system and, if this problem is fixable, this solution is preferable because the radio defaults to using the usbC port as the serial port so it should still work if a full reset is ever needed and the mechanism for usb should naturally accept as many different devices as practical because you never know what you will need to connect until you try it and it fails. As an amusing aside, the radio has a usb mode, but that's totally unrelated because in that case, USB stands for upper side band and is a very energy-efficient method of radio communication because it is only a spectrum of frequencies consisting of the sidebands of an audio signal without the carrier. It can be LSB or lower side band or USB for upper side band. Nothing like a few new acronyms for your day. I guess that's TMI. Martin McCormick WB5AGZ
Re: Proprietary USB Drivers; Ya Gotta' Love'em.
Gene Heskett writes: > You forgot to mention it can get the message thru better because it has a > 12 db advantage over competing noise compared to the original AM, > sometimes called Ancient Mary in our circles. I did forget that but you are correct. another interesting thing about ssb which includes both sidebands is that it was first tried during the 1930's as a concept and had to wait until after World War II before amateur radio operators, the millitary and commercial folks began using it routinely. To not stray too far from the group purpose, the usb dongles that give you a wide coverage receiver from 26 to almost 1000 MHZ use basically math to decode sideband, AM, FM and data using DSP techniques. Martin
Re: Proprietary USB Drivers; Ya Gotta' Love'em.
So, what is the easiest route to end up with a kernel that has this patch in it? The image for the current kernel is 4.19.0-5-686-pae The Raspberry Pi is an arm-based system and the image version number is probably different but the idea is the same, have a kernel that doesn't choke when one is using the ICR-30 usbC port. To put it bluntly, this is an area of Linux I have very little familiarity with because generally, Linux kernels just work so this is a left-handed complement in that I think this is the first kernel bug that I have encountered ever. I am sure I have run kernels with other bugs that I didn't know about but this is the first time one has bitten me, so to speak. Thanks. Martin WB5AGZ Martin McCormick WB5AGZ
eMail Com Between Hosts on a Private Net
The installation, here, is like millions of others. We are on a private VLAN with the router acting as the gateway to our ISP's network and the internet. There is nothing unusual about that so the question is Can systems on a 192.168.x VLAN use smtp to send mail to each other? I really should know the answer to this because I am a retired systems engineer who used unix systems and mailers all the time but that was in a university network and I would just assign a DNS name to each box and maybe a MX record if it was necessary so that box1.midlevel.edu could deliver mail to box2.midlevel.edu whether it was across the room or on another continent. The DNS support is what you don't have on a private VLAN so I want to do this in a safe but simple way. This would make it possible for Linux boxes on the network to send messages to the system I normally receive mail on so that squawks about a process crashing or some other problem are sure to be seen. The other systems sending those messages don't even necessarily need to send mail outside the network but they do need to send mail to the system I normally read mail on. I looked up this topic using duckduckgo and found very little hits that were on topic and lots of mercantile buzz about email hosting companies, etc. All are necessary but not what I was asking about. I have put static IP records in to the dhcp server on our router so 192.168.1.xx will always either have a specific host at that address or nothing if the MAC address changes and the record hasn't been updated. Also, I have put /etc/hosts files on Linux systems and a Mac and I believe there is a hosts file one can add to Windows systems for a similar effect. Thanks for any good ideas. Martin McCormick
Re: eMail Com Between Hosts on a Private Net
Joe writes: > Indeed so, it is even in a folder called 'etc', which I think is in a > 'drivers' folder. It's a while since I used it. There may be an > existing default or example \etc\hosts file. The LMHosts file is also > here, but of interest only to Windows networks for speeding up share > lookups. > > -- > Joe My thanks to all who replied and thanks for reminding me of some of the considerations one needs to think of when setting this all up since that's what I used to do as part of my job before retiring in 2015. As for there being 16 subnets in the 192.168.x.x number space, one can use variable-width subnet masks so unless you have a lot of hosts or, for some reason, are sharing space with another subnet, one can spread out to a Class B-style network which would be the whole space and have a subnet mask of 255.255.0.0 or you can make lots of little subnets of various mask sizes with blocks of 16 addresses being about the smallest subnet that is practical. Remember you always lose the all-zeros and all-ones broadcast addresses because they are special plus there is likely a router address taken by the router which leaves 13 addresses for hosts in each subnet. I remember early in the nineties our campus went from a single Class C to a Class B network and the whole campus used the Class B subnet mask with bridges holding the whole thing together. You should have seen some of the arp storms that would blow up like wild fire when somebody had a misconfigured host trying to be helpful or maybe deliberately misconfigured because somebody got their socks in a wad over something and started the fire on purpose. It wasn't long before we began more efficiently using the number space with big departments getting let's say 1024 or 512 possible addresses and others getting 16 or 32 addresses. By the time I left, we were using the huge private network space of 10.0.0.0 with aNAT or Network Address Translation to parts of the old Class B network and it generally worked well. I wrote a C then a perl program which would assign IP addresses for hosts on our networks that knew the sizes of all the subnet masks so assigning IP addresses was something anybody in our group could do without much more information other than what host name the customer wanted and what building it was in. My job was basically to encode all that in to rules for automation to keep us safe from making mistakes and it worked. Martin
Re: eMail Com Between Hosts on a Private Net
One more question I should know the answer to but am not sure of. The debian Buster system I use for email presently uses fetchmail to get mail from the ISP and is configured to use that ISP's smarthost for out-going mail. I do not want to effect (= muck this up) this functionality because it works well for now. Shouldn't I be able to install an imap server on the debian box and forward messages of interest to it, then reach imap4 on the private net from any system that speaks imap or has an imap client? That would do what I need to do. When I was researching, the article in wikipedia I read said that many commercial systems have email clients which understand imap, pop3, etc. The systems likely to do this on our network are a windows10 box, an iMac and maybe an iPad. The idea would be to forward an email message needing this attention to imap on the linux box, contact the Linux box from one of the devices I mentioned, and download the message at which point it would e as if that system had been hooked up to the ISP and received it. I was all ready to use .local as our domain name and then I looked that up and there is a good wikipedia article which explains how that is problematic and recommends using something like .lan, .office or something else that isn't likely to be registered as a top-level resolvable domain name. The machine I receive email on presently would be a good candidate for running a mdns but our netgear router advertises whatever dns's the isp uses for obvious reasons and that's fine but it would be nice if the mdns's address could also be known to clients on our network which could make DNS queries to each other's names that would resolve properly. Is there a way to advertise the mdns so that the router picks it up but doesn't drop the internet DNS's that we all need to resolve the rest of the world? I do remember when I was working, we explored open-source network authentication systems which involved fake DNS's that one had to advertise as such so their information wouldn't corrupt the proper working DNS's which could really mess things up if somebody happened to pickup and cache the wild card * that sent all new supplicants to the authentication server after they were already up and running. In our case, the corruption would be okay and done for good reasons but the dhcp server in our router already advertises two domain name servers so ours would have to be learned about by discovery. Thanks again. Martin McCormick
Curious Question about an Extra MAC Address
I was inventorying all the systems on our WiFi and wired network so I did the following: sudo nmap -sn 192.168.1.0/24 followed by sudo arp -a and I found nothing extraordinary so I also got on our Netgear router and told it to cough up a list of attached devices, showing the same list of stuff that the first two sweeps found. Additionally, there was 1 mac address with no other information such as an IP address or system name, just that MAC on a row all by itself. I had been looking for our DVD player which has the ability to view streamed movies and I think this might be it. As a computer user who is blind, I don't directly use this player but my wife does. We usually play DVD's from it however she did use it once to watch a movie off the network and it worked fine. I asked her if she remembered giving our WiFi password and she said no but was given a message to go to a certain url to authorize payment in order to watch the film and it all worked down to the monetary slurp so this is not a complaint or gripe about anything. My question is, did the link she went to on the computer or maybe her iPhone use javascript to get our WiFi password and if so, how did the player get the authorization to use our WiFi? The stray MAC showing up on the router is probably the player broadcasting an attempt to get in the network but to do so, it has to send the password in order to get a DHCP lease which would then allow it to get on the internet. I am just curious about how did this work and could a person cause it to log on to our WiFi network and receive a video stored on a system here? I'm not trying to rip anybody off but it's good to know what extra features are under the hood. What I am actually thinking of is more like a DVR where a computer can record a TV show like the old-school VCR's and then one could play it back through the DVD player on the network. This player can also update it's firmware when necessary and we've never been prompted from it for a WiFi password. It just seems to be able to get through when it needs to. I guess this could also be a neighbor trying to mooch WiFi, but I doubt it. Everything with an IP address belongs to us and there are no odd amounts of data coarsing through our router that aren't caused by us. This is a neighborhood in a college town so one must be careful but not too paranoid. Thanks for any good ideas. Martin
/usr/share/calendar.usholiday
There's a stale version of this file on my system that reflects when Debian was installed but it brings a question to mind. Does this file simply cause messages to appear on an event calendar or can it be used along with crontab in some way to either cause or prevent cron jobs from running if a date matches one of the dates in the file? I use cron and at jobs all the time and have thought that it would be nice to skip some jobs if the day is a holiday and the result might be not what one wanted. Right now, I don't think this file is doing a thing as it is for 2017 but this is why I am asking. I also saw versions of that file for other countries such as the UK. Thanks for all good responses. Martin McCormick
Re: /usr/share/calendar.usholiday
Greg Wooledge writes: > On Sun, Jan 16, 2022 at 09:58:11PM -0600, Martin McCormick wrote: > > There's a stale version of this file on my system that reflects > > when Debian was installed but it brings a question to mind. > > The one in the Subject: header? Are you sure that's the correct > path for this file? I goofed. I thought I could remember it long enough but it is actually /usr/share/calendar/calendar.usholiday I also still had an old directory tree from when I had wheezy on the system and there is /home/wheezy/usr/share/calendar/calendar.usholiday > > On Debian 11, I have the file /usr/share/calendar/calendar.usholiday I am presently running Debian GNU/Linux 10 as in Buster. --- > which is a different path from the one you showed. That was just a mistake on my part. Sorry to add more confusion. This also reminds me to get rid of the old wheezy files since I now don't even remember why I saved them but they are getting backed up nightly for no reason at all other than they get caught up in the sweep. Martin
Re: /usr/share/calendar.usholiday
deloptes writes: > And if you look into the file, you see it has no years associated with the > date, so it doesn't matter when it was written unless there was a change > in > the official holidays in the US. Correct. I see two types of date references, exact dates such as Christmas and New-year's and rule-based dates which are some designated day such as the second Monday in a given month, what the British refer to as a "bank holiday." My original question about this file has been answered in a round-about way in that there are also two types of holidays. St. Valentines day, for example, is not a postal or bank holiday but one will certainly get in deep trouble on the home or romantic front if that day is forgotten. Also days like Arbor Day are a nice idea but nothing about normal commerce or daily activities changes I also noticed that today in the United States should be marked but hasn't been yet. It is the third Monday in January which, since 1986 has been the official recognition of the birthday of a US civil rights leader, Martin Luther King Jr. The entry in that file probably should read: 01/MonThird Martin Luther King Birthday (3rd Monday of January) It is now a major holiday and schools and government offices are usually closed so it should probably be there. Thanks to all and sorry if I created any confusion about the path. Martin McCormick.
The Raspberry Pi that Took a Day Off.
This Pi is running Debian Stretch. I believe that's what version 9 is called. I have it capturing audio from a radio receiver and it's been doing that for several years now and it was doing that yesterday morning. Later in the day, I downloaded more audio and, after a long pause, I got the message from ssh that the Raspberry Pi wasn't there any longer so I retrieved the Pi from the room where it was and brought it to my Debian desktop system to work on it. I could login to the Pi which seemed to be up and running but the short story is that it couldn't talk to any address but our router.I couldn't even ping it's interface from the Pi, itself. If I was on the Raspberry Pi's console, everything looked normal as long as one wasn't using the TCP/IP interface. You could even do a ip addr or an ipconfig -a command and it would show that it had gotten the correct address from our dhcp server which is in the router. It would successfully ping the router but no other addresses, not even the address it uses on our network. I finally quit messing with it and went to bed but fired it up again today, January 25 and low and behold, it just came right up and is now back doing what it has been doing. Is it possible that it got a corrupted lease for dhcp from the router? Dhcp leases on our Netgear router are issued for not quite 24 hours so it may have gotten a bad lease, kept renewing it for the time it was powered up and then it got a new version of that lease today and all is well. It's the same IP address because I have put it in the router as a static IP address. The other thing that is weird is that the Raspberry Pi in question has both a wired Ethernet and a WiFi interface and both were misbehaving identically. Normally, the wired port is not used and the dhcp lease renewal process happens over WiFi. The results of ip addr always showed a correct subnet mask and the only rules in iptables are the 2 default rules. In other words, all looked normal except that it didn't work. While I had it on the work bench, so to speak, I ran fsck -fy on the SSD card since it has been a couple of years since I last did that and there was not a single squawk about anything. Thanks for any ideas about how things got wrong and then magically fixed themselves. When I turned it off last night, I gave it the halt -p command. The power supply has no switch so I unplugged it from power so it started fresh about 8 hours later. Martin WB5AGZ
fail2ban Squawk
I was attempting to setup a systemd timer and checking the syntax of that when I ran across a complaint from the fail2ban program which is a bit confusing. It reads: /lib/systemd/system/fail2ban.service:12: PIDFile= references path below legacy directory /var/run/, updating /var/run/fail2ban/fail2ban.pid → /run/fail2ban/fail2ban.pid; please update the unit file accordingly. So I looked in to that file and the actual line they were referring 2 is numbered 15 and points fail2ban.pid to /var/run/fail2ban/fail2ban.pid where it certainly lives with a recent date. What is the problem exactly? Enquiring minds want to know. Thanks for any and all constructive ideas. Martin McCormick
Is there any way to snoop on a USB port?
I have a Windows box that has software on it which programs two-way radios and it would be nice to know what the radio and computer are saying to each other. After trying a Windows application that reportedly can capture serial port traffic, I find that it doesn't appear to work with usb ports and unix/linux is my preferred world anyway so is there any sort of hardware that would pass through a USB connection from the Windows box to the radio and let me siphon off the traffic to a linux system and log it? The Windows app I tried to use has been around for a decade or more and probably works well with RS-232 ports but the traffic I need to grab comes from a usb device that creates /dev/ttyACM0 if plugged in to a Linux box and comm2 on the Windows system When I tried it today, it did nothing but complain that it was not connected. It finally dawned on me that it probably sees no serial ports at all as it is supposed to automatically find and log all port I/O and this is a relatively new HP Pavilion desktop computer which has no native RS-232 ports or parallel ports on it and, even if it did, the connection from the radio to the usb port is a usb plug with the usb hardware in it and a cable that plugs in to the radio that would most likely not be practical to tap. Thanks for any good thoughts. Martin WB5AGZ
Grub on Dead Disk
I had this same problem late last Fall on this same drive and am not sure what is going on because there is another Linux system here whose drive is arranged exactly the same way. An update in March modified grub and I didn't even give it a thought because that system just works but after an update a couple of days ago, the dead system starts a boot but grub complains about neither possible image being bootable. When I rescued that system last Fall, I was advised on this list to do the following which worked perfectly by the way: >If you want a grub-install command that writes /boot/grub files >somewhere onto /dev/sdd then you will first have to mount the desired >target boot partiton of /dev/sdd on some mountpoint that you choose, >and then run a command something like this: > sudo grub-install >--boot-directory=/some/mountpoint/where/is/the/sdd/boot /dev/sdd I mounted the drive on /mnt so looking at /mnt gives you / and the directory where the correct files are is then /mnt/boot or possibly /mnt/boot/grub. I created the following shell script which I named do_grub. #!/bin/sh sudo grub-install \ --boot-directory=/mnt/boot/ I get the following no matter where I point boot-directory: ./do_grub [sudo] password for martin: Installing for i386-pc platform. grub-install: error: install device isn't specified. I don't remember doing anything else. I think this time I may also need to do update-grub on the currently bootless drive. As with the last time this happened, the dead drive passes all fsck tests and the first time I ran fsck on it, nothing needed fixing. This is one of these problems that hardly ever comes up which means that when it happens, one spends/wastes a lot of time doing duckduckgo and puzzling. There should be a way to automate or at least prompt the person running the recovery attempt to find the right information to get things up and running if at all possible. It's all there such as the grub file in /etc/default which should have the uuid of the/ and swap partitions. Right then, one could see if the uuid's are correct and get a proper boot going in which one could then rerun update-grub if necessary. Right now, it's just fumble and fiddle and wonder what happened. I must have something set wrong to cause the apt-get upgrade to build a non-functional grub since the system in question made a number of reboots successfully between about last December 1 and now. I also remember reading a message about problems with swap which is on /dev/sda5 when the drive is booted. I found the messages log which shows swap as correctly mounted and working before the upgrade. Is there another log which might still have that message from apt-get upgrade. The message was not captured at the time. I have been using Debian on several systems for a number of years and this one system is the only one which seems to fall down every time grub is touched. That indicates I need to check one of the configuration files which probably has a wrong partition ID or something similar. I am just glad I have another Linux box to mount the no-boot drive on and that the drive, itself, does not seem corrupted just as the last time this happened. Thanks for any constructive ideas. Martin McCormick WB5AGZ
Grub on Dead Disk solved
I missed the INSTALL_DEVICE and now the script works. The next step is to see if the drive boots but I just forgot that one important detail. Sorry to waste anyone's time. Martin
Boot Repair
Is there any free utility that can run in Linux which helps one rebuild a corrupted boot configuration? I have a disk which is currently out of it's usual place as the boot drive for a debian system. The past two times that Buster updated grub, the drive became unbootable after the update. The drive, itself appears to be perfectly good and I think the problem which makes the update kill the drive is related to a dd copy I made some years ago of one drive to another so most of what is there is normal. After the second killing of the drive, I started to determine what is misguiding update-initramfs as grub-install seems to produce a grub.cfg file that is looking for all the right UUID's but when the system boots, grub now fails by trying to find a UUID that must have belonged to a now deceased system as it no longer appears in the fstab of either the newly-dead system or another system which is alive and well and being used to try to revive the injured drive. Everything needed to produce a valid boot is there if only I can run something that will reconstitude the boot image so that it looks for the right UUID's. I do remember seeing this message as the update ripped through the new grub: update-initramfs: Generating /boot/initrd.img-4.19.0-5-686-pae W: initramfs-tools configuration sets RESUME=UUID=6f6f4373-45a3-4c16-97b4-f62613 fe6d3b W: but no matching swap device is available. That UUID should start with a 3, not a 6 as the blkid for /swap is what should be there. I am not totally sure where tthe UUID that starts with 6f6f comes from but that's also the UUID one sees when grub crashes and burns on liftoff when trying to boot from that drive. Martin
Re: Boot Repair
Weaver writes: > On 03-06-2021 03:59, Martin McCormick wrote: > > Is there any free utility that can run in Linux which helps one > > rebuild a corrupted boot configuration? > > https://www.supergrubdisk.org/ Thank you very much as you did answer my question perfectly but I did not stop to think about all the possibilities of the form the solution might take. So, as a computer user who happens to be blind, is this stand-alone application using a linux kernel containing the speakup modules? On the Linux live CD's, one can get those to talk by striking the S key as the system boots up and then one can hear what is on the screen as one works. Do you know what kind of kernel it uses? Martin
Re: Boot Repair
Weaver writes: > https://www.supergrubdisk.org/ The site recommends downloading the hybrid version of grub2disk so that is what I did because all the Linux boxes I have are x86 hardware and have CD technology built in. This is quite an interesting creation. The ISO image is a wee bit over 16 MB in size and, as I suspected, the package includes no screen reader modules whatsoever but I did mount the CD after burning it and ran find . -name "*" -print and there are serial modules sprinkled throughout the distribution. There are occasional references to a serial console. In some of the documentation I have found on the site, I haven't found one iota of text as to how to make this console work but if computer users who are blind can access this serial console, that is like finding gold in them thare hills. Tomorrow, I will look further for any hints as to how to make the serial console appear but I actually did read the words "serial console" on that site. When I listed all the files on the CD, I also ran across the following: ./boot/grub/i386-efi/morse.mod ./boot/grub/i386-pc/morse.mod ./boot/grub/x86_64-efi/morse.mod If this has to do with outputting text in Morse Code, I know Morse since we amateur radio operators once had to know it to get our licenses and as long as it is sending at a reasonable rate which, for me, stops at about 20 words per minute, is perfectly useful to read the output. I didn't even see any references to Morse Code on the site so it's another way to make this useful. The PC being used to run this CD does have a piezo buzzer on the mother board so if there is a way to make it sing in Morse, I may not need the serial console but it certainly would be nice to know all there is about both of those fall-back methods. Thanks to anybody who knows where real sentences are that explain the serial console or the possible Morse code output. Martin WB5AGZ
Re: Boot Repair
Weaver writes: > Well, let us know how it goes, because I've noted a few visually > disadvantaged users on the list, and they would find the reference > useful. I have found out that grub is very accessible if one has defined a serial console and there is a working serial port on the target system. In this case, both conditions are true. The next thing one needs is a talking terminal which is no problem if one has another computer equipped with a serial terminal program like microcom or what I consider the gold standard, kermit which apparently has too many dependencies to be part of the Debian distributions any more. Kermit has a decent scripting language so one can save the commands that work, even the commands that involve entering a huge string in which one typo means completely starting over again. This means you can completely repeat all the things that work each time and feed it all in again in a fraction of a second. Talk about working smarter as opposed to harder. While I call kermit the gold standard, the gold standard for scripting is expect so one can call unix applications from the login shell to any valid command, read the output and generate commands as if one was there. You can even simulate human typing rhythms to defeat attempts to detect robots. I'd love to see a good CAPTCHA solver with OCR that was so good that CAPTCHA designers would go out and find honest work but I digress. This is actually a systemic problem these days when dealing with sick machines. Nobody is selling new desktops with RS-232 native serial ports and that is fine as long as there is an alternative way such as a bluetooth interface or ssh network connection. The dream solution would be a local network login covering all the lights-out conditions such as BIOS configuration and failure to boot as in this situation where grub is confused. Before I retired, some of our newest servers used management interfaces and ipmitool where one could turn power on and off, do BIOS setups and other configuration remotely. It was so cool to be able to do those things on a system in another town without leaving the chair or grabbing a go bag and blowing a whole day just to flip a couple of switches. Of course, if it was that simple, we usually could call someone we trusted at one of our remote campuses and ask them to flip the switches but I'm sure you see the advantage of remote management. For folks who are blind, you treat all those things as if they were halfway round the Earth instead of sitting less than 1 meter apart. They are all headless even if I can extend my arms and touch both machines. I think the grubdisk2 GUI app is basically a lost cause but that's the rule when dealing with accessibility. When it's done wrong, you chase down endless rabbit holes all day and muse as to how there must be some way to turn the sausage machine backwards and have live farm animals running away from the other end. There was an American country and Western song back in the early seventies which had a line at the end of each verse that went, "Work your fingers to the bone What do ya' get? Bony fingers." I think the best course right now is to take my bony fingers and workup a kermit script to interact with grub and get it to produce one boot and then use the running system and grub to fix itself. So far, nothing else is less work. I do appreciate your telling me about grubdisk2 because sometimes, these things turn out to be life savers and you have to try the rabbit holes to know for sure. Martin
Re: Boot Repair. Still Going Round and Round.
I placed the ailing drive back on a good Linux system and mounted it as /dev/sdd1 /mnt and ran the following commands on it: #!/bin/sh #mount the drive being repaired. Uncomment lines as needed. sudo mount /dev/sdd1 /mnt cd /mnt/boot #installing to the mounted disk sudo grub-install \ --boot-directory=/mnt/boot/ /dev/sda Installing for i386-pc platform. Installation finished. No error reported. root@wb5agz:/# But wait. There's more. root@wb5agz:/# update-grub --boot-directory=/dev/sda Unrecognized option `--boot-directory=/dev/sda' Not good. Continuing on to update initramfs. #Isolate the disk. sudo mount --bind /dev /mnt/dev sudo mount --bind /dev/pts /mnt/dev/pts sudo mount --bind /proc /mnt/proc sudo mount --bind /sys /mnt/sys That all worked. sudo chroot /mnt root@wb5agz:/# update-initramfs -c -k 4.19.0-5-686-pae That appears to work. root@wb5agz:/# update-grub Generating grub configuration file ... Found linux image: /boot/vmlinuz-4.19.0-5-686-pae Found initrd image: /boot/initrd.img-4.19.0-5-686-pae Found linux image: /boot/vmlinuz-4.9.0-9-686-pae Found initrd image: /boot/initrd.img-4.9.0-9-686-pae Now this ruins everything. Found Debian GNU/Linux 10 (buster) on /dev/sdc1 done This was a chroot environment but it appears that /dev/sdc1 was recognized as the boot source for this external drive. /dev is part of the system. The external drive normally lives in another system and is the only boot source which is why it should be /dev/sda1. Is there any way to run update-grub from grub itself? The system this external drive is meant for normally only has 1 boot drive which is the one that is currently confusing grub. If I could run install-grub and update-grub from the dead system, there would not be any other bootable drives and the only game in town would be /dev/sda for the boot record and /dev/sda1-5 for the boot drive. Understanding what is supposed to happen isn't particularly difficult but getting grub to think simple and easy Putting the drive on another system and hoping for no contamination seems to be a lot easier said than done. Martin
A Grub Boot Question about initrd
I have a plan but I need some more information. Is there any personalization done by the boot setup process? Do our UUID's or any other specific information pertaining to the installation make it in to the initrd files? While reading about the boot process, it doesn't appear that the initrd files or initranfs are personalized with anything pertaining to this computer and this installation. If that is so, then two computers using the same processor type should be able to use copies of the same initrd files and the only thing that is personalized on an individual computer is all the grub configuration in which the UUID's of at least / and /swap partitions are sprinkled throughout grub.cfg and /etc/default/grub. One should be able to write a program to get the appropriate UUID's out of fstab on the working system and translate them in to corresponding UUID's for the system on the operating table. If the target system actually boots, it is probably a good idea to run update-grub to make sure that still produces a working boot but this would still more than likely produce the same results if this process works in the first place. It's also possible that the reconstructed grub setup is okay except for the drive designation which usually starts out as /dev/sda1. On my good Buster system, this is now /dev/sdc1 and on the sick one, the attempt is being made on /dev/sda1. The idea here is to copy the concept of what is happening rather than the literal configurations which definitely will never work unless one used dd to generate the clone drive and I have actually done that once and it worked but the cloned system was then adapted for other uses. Here, all I want to do is rescue the boot process so it lives on and not have to reinstall everything else. As an aside, one ought to be able to do something like this. It makes life a lot simpler. Both systems are using the same kernel and versions of the same processor the only real differences are the UUID's. The grub configurations of both are the same down to the serial console. Martin
Re: A Grub Boot Question about initrd
Reco writes: > On Sat, Jun 05, 2021 at 12:46:13PM -0500, Martin McCormick wrote: > > I have a plan but I need some more information. Is there any > > personalization done by the boot setup process? > > Yes. One of the GRUB's tasks is to supply kernel which is about to boot > with root=... cmdline parameter. Root filesystem UUID can be used for > this. > > > > Do our UUID's or any other specific information pertaining to the > > installation make it in to the initrd files? > > In Debian - no, unless you include it there for some bizarre reason. > It's not needed for the things initrd usually does. > > > > If that is so, then two computers using the same > > processor type should be able to use copies of the same initrd files > > and the only thing that is personalized on an individual computer > > is all the grub configuration in which the UUID's of at least / > > and /swap partitions are sprinkled throughout grub.cfg and > > /etc/default/grub. > > It's not the CPU difference you need to worry about. > Different SATA controllers, video cards, NICs - i.e. what they call > periphery devices - those things require different kernel modules that > should be (or could be) used in early boot process, and therefore need > to be included in initrd. > > Luckily, Debian uses initramfs-tools for building initrd, and > initramfs-tools should build initrd with everything and a kitchen sink > included (MODULES=most in /etc/initramfs-tools/initramfs.conf). > > > > One should be able to write a program to get the > > appropriate UUID's out of fstab on the working system > > and translate them in to corresponding UUID's for the system on > > the operating table. > > Er, they've invented filesystem labels for exactly this many decades > ago. > > > > As an aside, one ought to be able to do something like > > this. It makes life a lot simpler. Both systems are using the > > same kernel and versions of the same processor the only real > > differences are the UUID's. > > Perfectly possible for the last 15 years or so. Assuming Debian and > MODULES=most, of course. > > Reco > > It sounds like I haven't missed anything obvious so I will see if I can write a perl script or some other text-muncher that will recreate the grub configuration of the working system but with the UUID's of the non-working system plus a grub.cfg file exactly like the grub.cfg file on the system that refuses to boot but with the correct UUID labels describing the boot partition on that system plus a copy of the kernel and it's module directory. As for partition labels, I have always thought the name was much easier to deal with than the 36 randomly-selected characters that make up the UUID on a unix partition but UUID labels are supposed to be unique and are what you encounter today so I will see if I can make the script modern compatible. Thanks to you and delop...@gmail.com.
Re: A Grub Boot Question about initrd
First I greatly appreciate all this information as the idea is to
fix a problem I probably created long ago though I am not sure
how but the short story is that apt-get upgrade ran update-grub
and update-initramfs late last Fall and I was able to rescue it
but it happened again at the end of April so I figured I had
better fix it correctly since I didn't know it was a ticking
bomb.
Felix Miata writes:
> Martin McCormick composed on 2021-06-05 12:46 (UTC-0500):
>
> > I have a plan but I need some more information. Is there any
> > personalization done by the boot setup process? Do our UUID's
> > or any other specific information pertaining to the installation
> > make it in to the initrd files?
>
> Dracut includes a root=UUID= statement in the initrd. A root=
> statement on
> the Grub linu* line overrides it. The override can be any of the accepted
> forms,
> including root=LABEL= and root=.
Good to know.
> The translation is easy. UUIDs are assigned to devices. The device names
> are
> acceptable substitutes, just as they used to be the only option before
> libata made
> device names capable of playing musical chairs. If devices also have
> labels,
> labels can be used instead of device names.
Very much like the reasoning behind udev which is a God-send if you
have multiple sound cards as one example.
> UUIDs were designed for computer scripts, not human brains. You'll need
> to make a
> conversion, but instead of UUIDs, I highly recommend using labels when
> you know
> device names will be different on the system migrated to. UUIDs are
> appropriate
> too, just rather unwieldy unless scripted.
There's 1 UUID for the root file system in /boot/grub/grub.cfg
which is the file one must not normally edit by hand. It
reads
menuentry 'Debian GNU/Linux' --class debian --class gnu-linux --class gnu
--class os $menuentry_id_option
'gnulinux-simple-658da8df-942e-4ad7-8c5b-695f2f2eb4ed' {
> Grub provides a shell. From the shell it's possible to boot using the
> information
> you have. Instead of trying to figure it all out in advance, drop to the
> shell and
> use the device names you know. It's really not hard if you have a copy of
> a grub
> menu from the two PCs planted in your brain or on paper. Once you've
> booted, it's
> easy to fix everything up by running grub-mkconfig, and if necessary,
> /etc/default/grub.
>
> > As an aside, one ought to be able to do something like
> > this. It makes life a lot simpler. Both systems are using the
> > same kernel and versions of the same processor the only real
> > differences are the UUID's. The grub configurations of both are
> > the same down to the serial console.
>
> Don't let the massive volume of data in grub.cfg mislead you. Its primary
> job is
> to find and load a kernel and initrd. If you know where they are and
> their names,
> none of the menu components or apparent search and identification bloat
> are
> necessary. Before Grub2 you could simply tell Grub's shell the root
> device, the
> kernel name (with associated parameters, most of which aren't actually
> necessary),
> and the initrd name, then boot. With Grub2 you may need to precede the
> root device
> name with loading a module or two or three, which you see in your copy of
> a
> grub.cfg, then proceed with the important three. I do a lot of cloning,
> so I've
> done it often.
Thanks again. I was starting to get discouraged but I
will try using actual labels since I thought you really shouldn't
do that.
Martin
Re: A Grub Boot Question about initrd
Thanks to all for the advice and knowledge you shared about how grub works. I am writing this on June 6 and early this morning, I edited the boot command in the grub shell after verifying that my stubborn no-boot drive truly was sitting at hd0,1msdos and grub-install had picked out hd2,1msdos for every single drive reference, even the bare metal one. I did try that one first in case I was missing some magic but the only drive on the whole system was that one and listing the contents of (hd0,1msdos) show exactly what you see if looking at / so grub showed it exactly where one would expect it to be and hd2,msdos is pure science fiction on that system so I set every drive reference to hd0,1msdos except the root which grub wanted to set to /dev/sdc1. The sample that Greg Wooledge showed looked very close to how my grub.cfg looked after doctoring the drive references to point to hd0,1msdos. and I tried booting that way and set the root parameter to /dev/sda1, /dev/sdb1 and /dev/sdc1 which is the one that grub chose, again non-existent. /dev/sdb1 is actually a Zipdisk drive but I tried it anyway since there is no zipdisk in there. It never booted so I am throwing in the towel, figuring that in all the messing around, I have corrupted something and wasted about two weeks on and off trying this, that and the other. I have a good 32 GB drive left over from a computer that died last year so I will just clone it from the working system that is sending this email. The drive that can't boot but is otherwise fine will now be the second drive and will mount as /home where all the important stuff I used it for can still live. The only good thing about this exercise is that I know a lot more about what grub does than I did and I also discovered that while I had perfect backups of the home directory, I had failed to set rsnapshot to backup /boot which was quite a stupid oversight on my part. /var, /etc and several other directories off of / get daily backups and I could have sworn /boot did but it's not there so I have been doing a lot of industrial-grade swearing lately. Grub is a very fine tool when used properly and I do believe that there should be a pristine way to make a disk bootable that is intuitive and automatic plus unaffected by the configuration of any other drive on the system, like the old sys command in DOS to make any DOS diskette bootable. The unix incarnation of the sys command would have to do a lot more such as verify that you have a kernel and modules ready to go plus consult your /etc/fstab and /etc/default/grub files but that shouldn't be a real problem. One would basically do the grub install in a a total jail so it would come out set as if one had just been running the installer and this was the only drive on the system. Thank you all for your help and sorry that it took so long to determine that nothing simple was going to work. Martin
Re: A Grub Boot Question about initrd
Felix Miata writes: > IMO you gave up too soon. IIRC you never showed us output from parted -l > or fdisk > -l. Very likely on the problem PC the / filesystem was/is not on the first > partition, where often lies a swap partition. Very likely root=/dev/sda2 > would > have been/be correct. Sorry. I meant to do that. Swap is the last partition. Disk /dev/sde: 29.8 GiB, 32017047552 bytes, 62533296 sectors Disk model: 600 Units: sectors of 1 * 512 = 512 bytes Sector size (logical/physical): 512 bytes / 512 bytes I/O size (minimum/optimal): 512 bytes / 512 bytes Disklabel type: dos Disk identifier: 0x0080 Device BootStart End Sectors Size Id Type /dev/sde1 2048 59314876 59312829 28.3G 83 Linux /dev/sde2 59314877 62533295 3218419 1.5G 5 Extended /dev/sde5 59316925 62533295 3216371 1.5G 82 Linux swap / Solaris $ ls -1 /mnt/ bin boot dev etc home initrd.img initrd.img.old lib lost+found media mnt opt proc root run sbin srv sys tmp usr var vmlinuz vmlinuz.old
Re: A Grub Boot Question about initrd
I admit I made several big mistakes, here. The first was not having a backup of /boot as I thought I did. The next is thinking I could just copy the whole boot directory from a known working system and be able to get it to work by using sed to replace the UUID's of the system it was on with those of the system it will be on. Had I backed up the broken system instead the minute I knew there was a problem, it would have been fixable. Trying to transplant the good boot system to a different machine really should work if one knows what all changed but if that was the case, I'd be crowing about how I got it working. I may try one last thing which is to decompress the initrd files and see if there is a way to make sure they work after transplantation. At this point, it's probably more important to know what is poisoning the well as a training exercise since having good backups of /boot would have turned a major halt in to a minor irritation. Trust me. The good system is going to get daily backups of boot and grub every 24 hours from this day on. I haven't actually done anything yet to clone the good system's boot drive because I want to do the clone in the middle of the night when things are quieter and I don't run fetchmail in the wee hours so the backup sees a more stable file system. If, by chance, I end up actually making the broken system work, I'll let you know what it was and the one drive can serve the whole system again. Martin
Re: A Grub Boot Question about initrd
Felix Miata writes: > This is a big lurking booby trap that could have been the problem both > last time > and this time. It's one of the reasons why installation systems and Grub > switched > from using device names to using UUIDs: inconsistent and/or unpredicable > device > enumeration. > > Is this a PC with both PATA and SATA controllers? These old PCs can > compound the > issue. > > Different kernel, disk controller, USB controller and BIOS combinations > can cause > e.g. an SATA or PATA disk sda to become sde, or sdb become sda, IOW, > device names > that don't persist. UUIDs are supposed to prevent inconsistent device > names from > being a problem, but in a rescue environment, it can muddy the waters, or > backfire. Yes on all counts. These are old PC's that use IDE controllers. The IDE cables end in the normal 34-conductor header with one pin missing on one side that plugs directly in to the drive. In this case, the connector plugs in to a SATA-to-IDE converter with a female SATA connector. I believe you have hit the nail on the head because I know for a fact that the devices change names. On the working system, the boot device we would normally think of as /dev/sda is /dev/sdc and 5 minutes from now might be /dev/sdq depending on what drives were plugged in to usb ports when the system booted. Just yesterday, I mounted /dev/sdd, thinking it was the now infamous unbootable drive and immediately realized that I was looking at the boot drive that starts the good system. I had plugged a thumb drive in to a usb port some time earlier and then rebooted the system while failing to remove the thumb drive as it wasn't even being used right then. Let's see. $ dd if=/dev/sdd of=/dev/sde What could possibly go wrong? Just kidding this time but that's what we are dealing with. That possibly lethal command can work but only under the right circumstances so if one forgot to recreate the world one was working with, it's tuition in the school of hard knocks. I haven't yet totally given up and done anything else rash so it's still full steam ahead. This is one of those situations where if a person has the time, this can be a teachable moment. I've got several thumb drives that I mount based on their UUID and that method of doing things is your friend if you understand how to use it. If not, what one ends up with is total chaos. Felix's message hasn't saved the day yet but it confirms my hunch that the UUID's have been at the heart of this problem all along because there is simply no other logical reason why one can not transplant the same boot logic from one system to the other when they are functional copies of each other. The trick will be to replace every UUID personalization which is doable. Whether it is practical still remains to be seen. To briefly reference an earlier message, I checked the good system to see what rsnapshot did over night and /boot is now committed to a series of little capacitors which is how SSD's remember things. From fstab: # /etc/fstab: static file system information. UUID="49c4eda8-3bcc-445e-841b-513b07d560fe" /var/cache/rsnapshot ext4 relatime,rw,user,noauto 0 0 If that drive wasn't always plugged in, the boot drive would have a different device number than /dev/sdc1 which is also mounted by UUID. That drive also plugs directly in to a usb slot and not any extender to hopefully insure that it is more likely to be present. I will uncompress the initrd files and see how hard it will be to find the UUID and change it to what it should be on the problem target system. Again, thanks to all. Martin WB5AGZ
Grub Success at Last
I did it! It works! Okay. Here's the short story. I read some more stuff about building a boot drive for another system than the one being used for the rescue. In this thread were the usual tales of woe which I have also experienced when misusing grub such as "Oh #%^*! Now I've got 2 systems that won't boot." a poster on that thread states: I'm not a grub2 expert (sorry) but try adding --skip-fs-probe to your grub-install line, I have found this prevents creation of /boot/grub/device.map which can cause booting to a grub prompt. I think that without this parameter grub-install, instead of doing what you tell it, thinks it is cleverer than you, and may do something different. Another thing is to be sure you are using the right grub-install (i.e. for grub2 and not for original grub). This isn't a problem if you are inside Centos but with SystemRecoveryCD both versions are available and so you have to use grub2-install. I learned the hard way... And as @wurtel pointed out (kudos), you should specify a drive not a partition. Grub2 installs in sector 0 of the whole disk drive, and this 'stub' is what runs at boot time, but it needs to know whereabouts on the disk it should install the files for the next stage of booting - this is what the --root-directory parameter is for. (I think.) Reading man grub-install and googling I see that --root-directory is not really meant to be used for grub2 versions 1.99++, though it does work in my experience. You are meant to use --boot-directory and refer to the actual boot directory, so this would give you: grub-install /dev/sdb --skip-fs-probe --boot-directory=/media/new_drive/boot End of quote I did the following: #!/bin/sh #mount the drive being repaired. Uncomment lines as needed. sudo mount /dev/sdd1 /mnt cd /mnt/boot #installing to the mounted disk sudo grub-install \ --skip-fs-probe \ --boot-directory=/mnt/boot/ /dev/sdd When you boot your disk that starts out on /dev/sda1, it's the first partition, in this case, so it will be /dev/sda1 not /dev/sdd1 . This was quite a long thread and people were remarking on how it is easy to get confused and do the wrong thing. Anyway, It booted right up and then I discovered there was no network. The network issue was only tangentially related because the update-upgrade process also install some more strict and modified ethernet interface naming rules and eth2 just doesn't cut it any more. Now, it's enp0s17 on the system that just got revived and enp0s18 or something similar on the system that got the same upgrades but is a little newer and took the tharapy much easier. Everything's working again at least in the Linux world. I saw this ad on TV for some product that is supposed to improve brain function and memory but I can never remember it's name. I could use a few more IQ points and things would flow a little faster around here. Thanks for all the help, everybody. Martin
SSD Memory Card (was The Raspberry Pi that Took a Day Off.)
I thanked the person who responded to my post and reported that there were no unusual log entries in syslog on the failing system so not much to go on. I decided to upgrade the Raspberry Pi which was suddenly having this mysterious problem as I have backups of the failing system so I figured I'd just clone a buster image from another Raspberry Pi that is running buster and use dd to copy it to the Raspberry Pi which was running stretch and having that weird problem. Then, assuming the upgraded Pi was alive, I could use the backup from the stretch system to put all the user space from my old home directory on the new buster system but I messed up again. The SSD card from the older system is 28 GB and the working buster system's SSD card is 32 GB so the smaller card filled up and I achieved that "no space left on device" state that we all love to see. Simple, I thought. I'll just delete all partitions, and do the dd if=rpi_good.img of=/dev/sdh which is where this card shows up when it is in the reader and I can maybe dd a smaller 8.5-gb drive which I have that has almost never been used. Murphy has struck again. Here is what happens when I try to use that 28-gb SSD card: Script dump follows Script started on 2022-01-26 20:03:52-06:00 [TERM="Linux" TTY="/dev/pts/3" COLUMNS="80" LINES="25"] 1wb5agz martin tmp $ sudo su - [sudo] password for martin: root@wb5agz:~# fdisk -l /dev/sdh /dev/sdh: 28.8 GiB, 30908350464 bytes, 60367872 sectors Disk model: USB HS-SD Card Units: sectors of 1 * 512 = 512 bytes Sector size (logical/physical): 512 bytes / 512 bytes I/O size (minimum/optimal): 512 bytes / 512 bytes Disklabel type: dos Disk identifier: 0x680226ff Device[0m [1mBoot[0m [1m Start[0m [1m End[0m [1m Sectors[0m [1m Size[0m [1mId[0m [1mType[0m /dev/sdh1 8192 137215 129024 63M c W95 FAT32 (LBA) /dev/sdh2 137216 62333951 62196736 29.7G 83 Linux root@wb5agz:~# fdisk /dev/sdh Welcome to fdisk (util-linux 2.33.1). Changes will remain in memory only, until you decide to write them. Be careful before using the write command. Command (m for help): p Disk /dev/sdh: 28.8 GiB, 30908350464 bytes, 60367872 sectors Disk model: USB HS-SD Card Units: sectors of 1 * 512 = 512 bytes Sector size (logical/physical): 512 bytes / 512 bytes I/O size (minimum/optimal): 512 bytes / 512 bytes Disklabel type: dos Disk identifier: 0x680226ff /dev/sdh1 8192 137215 129024 63M c W95 FAT32 (LBA) /dev/sdh2 137216 62333951 62196736 29.7G 83 Linux Command (m for help): d Partition number (1,2, default 2): Partition 2 has been deleted. Command (m for help): d Selected partition 1 Partition 1 has been deleted. Command (m for help): w fdisk: failed to write disklabel: Operation not permitted root@wb5agz:~# logout 2wb5agz martin tmp $ exit Script done on 2022-01-26 20:07:59-06:00 [COMMAND_EXIT_CODE="1"] So close and yet so far away. If I try to use dd on /dev/sdh again, I get: 1wb5agz martin tmp $ sudo dd if=~/rpi/rpi2_good.img of=/dev/sdh [sudo] password for martin: dd: writing to '/dev/sdh': Operation not permitted 1+0 records in 0+0 records out 0 bytes copied, 0.021666 s, 0.0 kB/s 2wb5agz martin tmp $ exit I've had usb thumb drives fail when filled to capacity and, if I hadn't been tired, I would have noticed that the image was 4 gb too large so this wouldn't have happened but surely, there is some way to at least zero out everything so the SSD can be either reformatted or used as the output of an image transfer using dd. The SSD passed a fsck test earlier in the day before I blew it up so the chip should be salvageable. I don't care for recovering either of the two partitions which will be overwritten anyway if the SSD can be made writable again. One can mount /dev/sdh1 which is the fat32 partition but it mounts as read-only. If you try /dev/sdh2, you get a squawk that says that this partition is not in /etc/fstab which is certainly true but it doesn't mount either so mount doesn't understand what I was trying to do. This is the partition that got the overflow condition so it's utterly useless. These last 2 days, I have been a day late, a Dollar short and 1 step behind on what shouldn't be really that big of a problem. Martin WB5AGZ
Re: SSD Memory Card (was The Raspberry Pi that Took a Day Off.)
Charles Curley writes: > I'm no expert on RPis, but that sounds to me like the SD card is > protected against writes. Check for any physical write protection > switches on the card itself and the holder. Thanks for the suggestion, but this is one of those SSD cards that often is found in a camera and resembles a wafer the size of a thumbnail. It has a projection that acts as a key way to keep a user from inserting it in the wrong way and there is a groove for a fingernail to help pull the chip out of the socket. This particular one was reading and writing just fine until I bricked it by the DD that must have overwritten some address which now makes it unwritable. It went from good to bad without my removing it from the card reader so there should be some way to at least clear it for writing again. Apparently, it stops being writable if the partition table is corrupted. In my case, I just want to delete both partitions and start over. The OS is debian Buster and has all the tools you can expect to find and runs on a 64-bit ARM. Otherwise, it's pure debian Linux. Martin
Re: SSD Memory Card (was The Raspberry Pi that Took a Day Off.)
Bijan Soleymani writes: > Can you delete both partitions, create a new single linux partition, > reboot > then run mkfs.ext4 to create a single new partition and then just install > linux onto it or try dd again? Great suggestions but I can't. Part of the typescript output I included was me doing just that and I was root when I did it but the squawk is that I don't have permission as if I wasn't root. If writing to the SSD card was possible, I could have just done the dd command, only with a smaller input image and it would have worked. Thank you because what you suggest would be the normal way out. Martin Martin
Re: SSD Memory Card (was The Raspberry Pi that Took a Day Off.)
David Wright writes: > I've not heard of that problem. You were prevented from zeroing the > entire device, which would have wiped the partition table anyway. > > What I would want to check is that the OS isn't doing something > stupid, like trying to automount it, failing, and consequently > setting the device readonly. By OS, I really mean DEs, or > automounters in general. > > You could also try zeroing it in another machine, ± any adapters > required. (Bear in mind that adapters do have readonly sliders.) I suspect this is the crux of the problem. the adapter I connected is a card reader. You put the SSD in a little plastic jacket that holds the SSD in such a way that the card reader can access the edge connector but the holder jacket has no electronics. There is a small notch in the plastic of the jacket on the left edge and the right front corner of the plastic carrier has a diagonal cut to prevent someone from putting it in upsidedown. Since I posted, there is good news but I still wonder if I am not going bonkers because after unplugging the Sony card reader and plugging it back in, I now am getting device /dev/sdg instead of /dev/sdh. I was also able to do the following: #sudo fdisk /dev/sdg which gave me the fdisk utility as before so I did what crazy people do which is to do the same thing as before, hoping for different results. By Joe, I got them. I typed d to delete a partition and it put partition 2 up as the default candidate as before so I selected it and then typed d again which told me that only partition 1 was left so it was deleted. I had gotten this far before so wasn't too excited but type w and this time got the message stating that the partition table had been rewritten and fdisk then exited. Now, doing sudo fdisk -l /dev/sdg yields 1wb5agz martin tmp $ sudo fdisk -l /dev/sdg Disk /dev/sdg: 28.8 GiB, 30908350464 bytes, 60367872 sectors Disk model: USB HS-SD Card Units: sectors of 1 * 512 = 512 bytes Sector size (logical/physical): 512 bytes / 512 bytes I/O size (minimum/optimal): 512 bytes / 512 bytes Disklabel type: dos Disk identifier: 0x680226ff The partitions are gone. My latest screwball theory is that the Sony card reader went in to some sort of protective mode after the dd operation overwrote the device. My unplugging the reader and plugging it back in reran the driver which reset the protective mode back to normal which may be why it all worked right this time. One last question: Since the image will still be too large as it is, can tunefs be run on it or a copy of it to shrink about 4 gb of user space? The good system I copied the image from only had about 12% of the partition used so I should be able to transplant it to the smaller disk if tunefs can do that and still leave a bootable device. Thanks for all useful ideas. Martin McCormick
Re: SD Memory Card (was The Raspberry Pi that Took a Day Off.)
My thanks to everybody who has responded here. I think the prudent thing to do is use a new SSD card and I have one that is supposed to be a full 32 gb. The card I was able to finally clear the partitions on is several years old but I think it is still good but the suggestion to use it in a less mission critical application is a good idea plus it is not a full 32 gb. I have a radio receiver that can store information on one of these cards and I'll use it for that since if there turns out to be a bad cell somewhere, the world won't end like it can if one's kernel or sector information becomes corrupted if it happens to land in a vital piece of code. Again, many thanks. Martin McCormick
Re: SD Memory Card (was The Raspberry Pi that Took a Day Off.)
There's always one more question that nobody mentions and none of the articles one finds on the topic don't touch. When looking at the man page for resize2fs in debian, it talks about the -b option to turn on "the 64 bit feature." __ When shrinking the size of the partition, make sure you do not make it smaller than the new size of the ext2 filesystem! The -b and -s options enable and disable the 64bit feature, respect- tively. The resize2fs program will, of course, take care of resizing the block group descriptors and moving other data blocks out of the way, as needed. It is not possible to resize the filesystem concurrent with changing the 64bit status. __ Do I need this since the Pi runs an ARM processor which would make the command #sudo resize2fs -b /dev/loop0p2 6.3g or is it the same command without the -b flag? I have determined that the 28.9 gb SSD card is 10% full with the debian installation and the files I want in my login directory. I found a 7.3 gb SSD card that has probably never been used that came with the first Raspberry Pi I bought around 2012 or so, going for a 32-gb card instead. If I shrink the Linux partition to 6.3 gb which is what the small card had available, I should have it about 40 or 50% full. I can then safely dd it on to a larger card any time I want to do so and then use resize2fs to expand the Linux partition after it is installed. What I did so far was to mount the 27-gb partition on /mnt through /dev/loop0 and edit /mnt/etc/hostname to reflect the host name for the system being rebuilt. The edit changes the image which is really neat. All that is left is to shrink it down to 6.3 gb and it should be ready to dd on to the 7.3 gb card which should be bootable on it's own but which I will use to seed a new 28-gb system that can be customized after it is running. Many of the raspbian distributions have a #1 partition that is a small fat32 lba partition for Windows users to be able to activate debian from Windows. Is this even necessary once one is using unix tools on the disk? Thanks to all the good advice from everyone. I am seeing the end of this project and have learned some new useful tricks that are good to know. Martin
Resize2fs Questions
I figured I would start a new topic as none of this pertains to the previous messages I posted. I've got an almost 30-GB disk image of a working debian installation for a Raspberry pi that I should be able to easily squeeze on to a roughly 8 GB SSD card because it only takes up 10% of the 30-GB card and I think I am close but not there yet. Because I hate drudgery with regards to syntax, I wrote a little shell script that makes it easier to duplicate the same actions and also makes it easy to show exactly what is happening so far. The big image name is rpi1_good.img and it is 29 gb in size. Here are the commands that work so far. #!/bin/sh #This resets everything for a new run. sudo losetup -d /dev/loop0 #copy a backup to the image name to start fresh. sudo cp -p rpi1_good.img.bak rpi1_good.img #This takes about half an hour. Now, get to business. sudo losetup -P /dev/loop0 rpi1_good.img #This should show 2 partitions on loop0 when successful. sudo ls -l /dev/loop0* #You get /dev/loop0p1 and p2 #e2fs is required to run before one can resize. sudo e2fsck -f /dev/loop0p2 #I should be telling resize2fs to squeeze everything in to a 7GB #partition. sudo resize2fs /dev/loop0p2 +7G There are no complaints up to this point and resize appeared to work so now it is time to reduce the size of partition 2. This may be where things go off the track. I run sudo fdisk /dev/loop0 and fdisk tells me it is 29.9 GB which I expected since Partition #2 should be mostly empty but for the OS and my login directory. I note that the starting sector for P2 is 137215 since one must start at that sector and then I delete P2 and then add a new partition which defaults to 2. fdisk prompts for the first sector with a default of 2048 but I type in 137215. The last sector is in the 62-million range and I want only 7 GB as the size so I type +7G and it takes that. I then type w to save the changes and there is no complaint. The first hint of trouble is when you do fdisk -l /dev/loop0. The size count is still almost 30 GB but then a ray of Sun in that P2 has a size of around 7 GB. I outwitted Murphy this time by using another file on my computer as the target rather than the SSD card so that if it went over 8 GB, nothing annoying would happen. I watched on another console as the size climbed past 5 GB, then 6 and 7 GB and then it was up past ten so it was going to do the whole 30 GB if I hadn't stopped it. I am not sure what I missed but the manual page for resize2fs warns you to use something like fdisk to trim off the excess empty space which is what I thought I did. I would also have thought that the disk image file should have deflated like a punctured air mattress but it didn't. I think I am almost there but obviously not yet. Thanks in advance for ideas. Martin
Fixing a Grub Foul-up
I have goofed, I think. There is a serca-2000-vintage Dell Optiplex that has been working fine up to yesterday when I did the usual apt-get update followed by the apt-get upgrade on buster. The update and upgrade appeared to work. One of the things that got visited was grub and it was then that I was reminded that there was another drive in the system that had a bootable image of buster on it also. Grub reported seeing it on /dev/sdc which is coorrect. This particular system has a zip drive that always shows up as /dev/sdb so the next hard drive after /dev/sda is /dev/sdc. I rebooted to make sure all was well and waited and waited . . . The system sits there like a bump on a log. I have a usb device that lets one mount IDE and SATA drives that are outside the system so I pulled the sata drive which is the boot drive for the now dead system and plugged it in to the usb converter. the drive breezes through fsck and looks perfectly normal. I looked at /boot/grub/grub.cfg which one is not supposed to edit as grub builds it based on /etc/default/grub which one does edit. If I was to mount that partition on a working system, it, of course, will have a different device number such as /dev/sde1 instead of /dev/sda1 which it should have when booting up the system it normally runs in. Is there a safe way to mount this drive, possibly using chroot, re-run grub-config and get the drive bootable again? If I look at grub.cfg and /etc/default/grub, everything looks as if it should work but it doesn't. I think boot problems are some of the most agrevating issues. They are true show stoppers. I've got backups but that's beside the point. Unless I can fix whatever happened, it's going to be quite a time waster. Thanks for any constructive suggestions. Martin McCormick
Re: Fixing a Grub Foul-up
Dan Ritter writes:
> Here's what you can do:
>
> On a good system, mount your drive. Let's pretend that it's
> recognized as /dev/sdg, and you have a /boot on /dev/sdg1 and
> a root partition on /dev/sdg2.
>
> ls -al /dev/disk/by-partuuid/| grep sdg
>
> will get you the partition UUIDs for that disk. One of them will
> be for /dev/sdg1 and another for /dev/sdg2.
>
> The kernel really likes these as root filesystems identifiers.
> The kernel parameter that you put in /etc/default/grub is
>
> ROOT=PARTUUID=dddf0dd6-dd6b-d542-9eac-015a765cd6f6
>
> although you will want to substitute in the appropriate
> part-uuid for /dev/sdg2.
>
> Finally, you can run
>
> grub-install /dev/sdg
>
> to get a new copy of grub into the master boot sector of the
> disk.
>
> Hope that helps,
It certainly does!
First, thanks to all who answered so far. I think you have
perfectly described what is causing the problem although grub is
not high in my knowledge base like it probably should be.
When I read /boot/grub/grub.cfg, I see
references to a uuid that isn't any of the partitions that have
UUID's. On my disk, Partition 1 is / and blkid spits out a
number that starts with 9F. Partition 2 was created to get an
extended partition of 5 which is swap space so 1 and 5 have
UUID's. The UUID referenced in grub.cfg starts with 3.
The rest of this message consists
of grub.cfg.
Martin
#
# DO NOT EDIT THIS FILE
#
# It is automatically generated by grub-mkconfig using templates
# from /etc/grub.d and settings from /etc/default/grub
#
### BEGIN /etc/grub.d/00_header ###
if [ -s $prefix/grubenv ]; then
set have_grubenv=true
load_env
fi
if [ "${next_entry}" ] ; then
set default="${next_entry}"
set next_entry=
save_env next_entry
set boot_once=true
else
set default="0"
fi
if [ x"${feature_menuentry_id}" = xy ]; then
menuentry_id_option="--id"
else
menuentry_id_option=""
fi
export menuentry_id_option
if [ "${prev_saved_entry}" ]; then
set saved_entry="${prev_saved_entry}"
save_env saved_entry
set prev_saved_entry=
save_env prev_saved_entry
set boot_once=true
fi
function savedefault {
if [ -z "${boot_once}" ]; then
saved_entry="${chosen}"
save_env saved_entry
fi
}
function load_video {
if [ x$feature_all_video_module = xy ]; then
insmod all_video
else
insmod efi_gop
insmod efi_uga
insmod ieee1275_fb
insmod vbe
insmod vga
insmod video_bochs
insmod video_cirrus
fi
}
serial --speed=115200 --unit=1 --word=8 --parity=no --stop=1
terminal_input serial
terminal_output serial
if [ "${recordfail}" = 1 ] ; then
set timeout=30
else
if [ x$feature_timeout_style = xy ] ; then
set timeout_style=menu
set timeout=5
# Fallback normal timeout code in case the timeout_style feature is
# unavailable.
else
set timeout=5
fi
fi
play 480 440 1
### END /etc/grub.d/00_header ###
### BEGIN /etc/grub.d/05_debian_theme ###
set menu_color_normal=cyan/blue
set menu_color_highlight=white/blue
### END /etc/grub.d/05_debian_theme ###
### BEGIN /etc/grub.d/10_linux ###
function gfxmode {
set gfxpayload="${1}"
}
set linux_gfx_mode=
export linux_gfx_mode
menuentry 'Debian GNU/Linux' --class debian --class gnu-linux --class gnu
--class os $menuentry_id_option
'gnulinux-simple-3584c851-6616-464f-8a8e-2d634fca5969' {
load_video
insmod gzio
if [ x$grub_platform = xxen ]; then insmod xzio; insmod lzopio; fi
insmod part_msdos
insmod ext2
set root='hd1,msdos1'
if [ x$feature_platform_search_hint = xy ]; then
search --no-floppy --fs-uuid --set=root --hint-bios=hd1,msdos1
--hint-efi=hd1,msdos1 --hint-baremetal=ahci1,msdos1
3584c851-6616-464f-8a8e-2d634fca5969
else
search --no-floppy --fs-uuid --set=root
3584c851-6616-464f-8a8e-2d634fca5969
fi
echo'Loading Linux 4.19.0-5-686-pae ...'
linux /boot/vmlinuz-4.19.0-5-686-pae
root=UUID=3584c851-6616-464f-8a8e-2d634fca5969 ro console=tty0
console=ttyS0,115200n8 quiet
echo'Loading initial ramdisk ...'
initrd /boot/initrd.img-4.19.0-5-686-pae
}
submenu 'Advanced options for Debian GNU/Linux' $menuentry_id_option
'gnulinux-advanced-3584c851-6616-464f-8a8e-2d634fca5969' {
menuentry 'Debian GNU/Linux, with Linux 4.19.0-5-686-pae' --class
debian --class gnu-linux --class gnu --class os $menuentry_id_option
'gnulinux-4.19.0-5-686-pae-advanced-3584c851-6616-464f-8a8e-2d634fca5969' {
load_video
insmod gzio
if [ x$grub_platform = xxen ]; then insmod xzio; insmod lzopio;
fi
insmod part_msdos
insmod ext2
set root='hd1,msdos1'
if [ x$feature_platform_search_hint = xy ]; then
search --no-floppy --fs-uuid --set=root
--hint-bios=hd1,msdos1 --hint-efi=hd1,msdos1 --hint-baremetal=ahci1,msdos1
3584c851-6616-464f-8a8e-2d634fca5969
e
Fixing a Grub Foul-up, Not There Yet.
I did some duckduckgo-ing about grub rescue and found useful
things but am still dead in the water.
I appear to be using grub, not grub2. One of the
articles I found had an example of how to use grub rescue that all
works except, of course, for the actual booting of the kernel.
There's an extra little wrinkle in that, as a computer
warier who happens to be blind, grub isn't really accessible in
that you can not get direct access to it via a serial port and
terminal, say, from another computer that is actually working. A
really clumsy work-around does work off of the dead computer in
that it is old enough that the Print-screen key dumps an ASCII
character dump of the screen to the parallel printer output every
time you press Print-screen. I found out after connecting a
second computer to a parallel-to-serial converter that I was in
grub rescue> mode.
Each time one types on a keyboard attached to the dead
computer, one then must hit Print-screen to find out whether it
goes or blows.
This is slow and awkward because there is no
character-by-character feedback so if you hit the n instead of
the m, you don't find out that it's a no go until you read the
whole screen.
Also the Tabs you hit don't show up in the output so
it'salloneunbrokenline.
Is there a way I can write the commands in to a text file
that I can call to at least know what commands I am executing in
grub rescue and do in a tenth of a second what it takes minutes
to do otherwise?
What I have right now is a text file called stuffcommands
and it is sitting in the boot level of the dead boot disk.
When I had the drive mounted on a working system, I did chmod +x
stuffcommands but since there are no shells running, I just put
the following lines in the file. Everyone of these commands
works when typed in except for the final one which boots the
kernel.
What I need to type in with, of course, no mistakes looks
like:
set root=(hd0,1)
linux (hd0,1)/boot/vmlinuz-4.19.0-5-686-pae root=/dev/sda1
#I have yet to see this one work.
initrd (hd0,1)/boot/initrd.img-4.19.0-5-686-pae boot
Here is a screen capture of what pressing Print-screen
produces The first line is that error message and is probably
the clue as to what went wrong. The successful commands were
laboriously typed in by hand once again to prove they work
although by now, I know the kernel version number about as well
as my last name.
error: symbol `grub_calloc' not found.
That's the smoke that the booster is going to blow up on the
launch pad.
Entering rescue mode...
grub rescue> ls
(hd0) (hd0,msdos5) (hd0,msdos1) (fd0) (fd1)
The root partition is Partition #1.
grub rescue> (hd0,1)/boot/stuffcommands
Unknown command `(hd0,1)/boot/stuffcommands'.
Ah, if only. Those few lines would make each retry so much more
efficient.
grub rescue> set root=/de/sda1
It worked.
grub rescue> linux(hd0,1)/boot/mlinuz-4.19.0-5-686-paeroot=/de/sda1
And that one worked!
grub rescue>
That much works and it would be so nice not to have to repeat
those steps every time. Is there a way to call that
stuffcommands text file so that grub thinks it's coming from the
keyboard?
I think I'd like a keyboard that one can store commands
like this in so as to keep the donkey work down to the problem
lines.
Actually, if only grub could be made to operate remotely,
I'd just script all the stuff that works and also know if I
fat-fingered anything when it happens so one could backspace and
fix it the way we fix any typing error.
Before I retired in 2015, I occasionally had to setup new
servers and these had lights-out minicomputers that helped one
remotely access even a powered-down box in another state,
province or country. The linux kernel, itself can easily be
configured to be accessible via a serial port or on the local
console so grub sorely needs headless capabilities for those rare
times where stuff just doesn't come back from a reboot.
I appreciate the good suggestions I have gotten from
several of you so far.
Martin McCormick
Re: 780 files in /usr/share/zoneinfo/
writes: > Suppose a hacker logs into your computer from far, far away, say > from somewhere in Nepal. > > Surely you'd want this person to see the time adapted to their > locale? That's the least courtesy you can be expected to provide? > > ;-P > > Now putting my tongue out of my cheek again: in Unix, a computer > "has" no time zone or language -- people have those. And since, > again in Unix, several people can be logged in [1] at the same time, > it's up to the user's environment [2] to decide on time zone, > language, etc. > > This concept is surprising at first coming from other cultures, > where Microsoft was happy to sell you another complete version > of Windows if you wanted your computer to talk to you, e.g. > Portuguese (and yet another for Brazilian Portuguese, greedy as > they are). > > Of course, Microsoft has caught up (they are trying since the > mid-90s), but not without some spectacular messups. Remember that > one where (I think it was Windows 95), while trying to automate > the spring DST transition were spotted dithering endlessly between > 2AM and 3AM? > > Unix has had this abstraction always: there's the internal time, > and there's the time shown to the user, which depends on the user. > > There's the error itself, and there's the error message shown to > the user. And so on. > > Cheers I just cd'd to that directory and it looks like there's about 1 GB there. I think it is cool to have all that info and I even configured a linux system for BST so cron would record shows I was interested in from the BBC. It works perfectly when we go from Summer time to standard time in Winter because North America and the UK don't switch at the same time but it all works. I've wished one could just set certain parts of the computer to other times but I can also understand why this could be a problem. By the way, I have never been outside the United States but am an amateur radio operator and learned at a very young age to appreciate those time zones if one wants to know when to listen for interesting things. Martin
Re: Fixing a Grub Foul-up, Not There Yet.
Felix Miata writes: > Save yourself many keystrokes by using the symlinks in the root directory > instead > of the long-winded full version-named /boot/vmlinuz-4.19.0-5-686-pae This is wonderful to know and in the root or / directory of this disk, there is initrd.img, initrd.img.old, vmlinuz and vmlinuz.old Those all point to valid targets. I presently have the no-boot disk mounted on a working linux system where it's root directory is /dev/sde1 and the current links plus the old links are as follows: lrwxrwxrwx 1 root root32 Jun 26 2019 initrd.img -> boot/initrd.img-4.19.0-5-686-pae lrwxrwxrwx 1 root root31 Jun 26 2019 initrd.img.old -> boot/initrd.img-4.9.0-9-686-pae lrwxrwxrwx 1 root root29 Jun 26 2019 vmlinuz -> boot/vmlinuz-4.19.0-5-686-pae lrwxrwxrwx 1 root root28 Jun 26 2019 vmlinuz.old -> boot/vmlinuz-4.9.0-9-686-pae I finally have found something that may be a clue to what's actually wrong. While trying to manually make that drive boot, I got this: grub rescue> set prefix=(hd0,1)//boot/grub grub rescue> insmodnormal Unknown command `insmodnormal'. That was a good old syntax error so I tried insmod normal with a space grub rescue> insmod normal error: symbol `grub_calloc' not found. I'm pretty sure that shouldn't happen at all and is what's behind the failure to boot. I haven't found any uuid's that are different although I first thought I had as I looked at some links which had uuid's but they were good when I looked at the actual partition. It's easy to go down a rabbit hole if one doesn't watch out. I think there may be something about grub that got left out or changed during the upgrade. Martin Martin
Re: 780 files in /usr/share/zoneinfo/
Andy Smith writes:
> Hi Martin,
>
> Are you sure about this? There is no Debian or Ubuntu host I have
> access to that has a /usr/share/zoneinfo/ that contains more than
> 4MiB of data. For yours to have 256 times this much is quite an
> aberration. What did you type to determine that your
> /usr/share/zoneinfo/ has 1GiB of data in it?
find . -name "*" -exec ls -l {} \; \
|grep -F / \
| awk ' { total += $5 } END { print total }'
That usually just adds the sizes of all the files it can
find all the way through the tree.
If this is not an accurate way to determine how many
bytes there are in a directory then that would be the reason for
the discrepancy.
>
> > I've wished one could just set certain parts of the
> > computer to other times but I can also understand why this could
> > be a problem.
>
> You can set any process to have a different time zone by use of
> environment variables.
>
> $ date
> Sat 21 Nov 21:34:55 UTC 2020
> $ TZ=America/Los_Angeles date
> Sat 21 Nov 13:35:04 PST 2020
That is very true but cron only works in the time zone
for wherever the TZ for the system is set.
On the system I have set to British time, my local login
shell has TZ=America/Chicago so I read Central Standard Time for
file time stamps but if I set a cron job, it runs on either
British Summer Time or British standard time which is the same as
UTC during Winter.
> Cheers,
> Andy
Re: Fixing a Grub Foul-up, Not There Yet.
Dan Ritter writes: > in /boot/grub/menu.lst > > serial --unit=0 --speed=9600 --word=8 --parity=no --stop=1 > terminal serial > > (yes, that's two lines) > > I hope that helps. > > If you want the option of either serial or console access, > replace the second line with > > terminal --timeout=10 serial console > > The default will be serial, but if you press a key on the > console within 10 seconds, it will go there instead. > > -dsr- Thank you. That is a tremendous thing to know. I thought there was something serial in grub but I couldn't remember. This is the first time I ever had any real trouble with grub and I can put this drive back on the working linux box and modify that file, assuming the change would take place on the next run of grub. Martin
Re: Fixing a Grub Foul-up
Dan Ritter writes: > Here's what you can do: > > On a good system, mount your drive. Let's pretend that it's > recognized as /dev/sdg, and you have a /boot on /dev/sdg1 and > a root partition on /dev/sdg2. > > ls -al /dev/disk/by-partuuid/| grep sdg > > will get you the partition UUIDs for that disk. One of them will > be for /dev/sdg1 and another for /dev/sdg2. > > The kernel really likes these as root filesystems identifiers. > The kernel parameter that you put in /etc/default/grub is > > ROOT=PARTUUID=dddf0dd6-dd6b-d542-9eac-015a765cd6f6 > > although you will want to substitute in the appropriate > part-uuid for /dev/sdg2. > > Finally, you can run > > grub-install /dev/sdg > > to get a new copy of grub into the master boot sector of the > disk. > > Hope that helps, > > -dsr- This does help a lot and I have read similar examples how to work on a drive that has been mounted on a different system than the system on which it will ultimately be used but I am still doing something wrong and the results are dangerous to say the least. At one time, the otherwise good drive was mounted on /dev/sdd with the root partition on /dev/sdd1. I typed sudo grub-install /dev/sdd. It ran for a few seconds, announced that grub was installed without any errors and exited. After looking at /dev/sdd1/grub and seeing no updated date stamps, I had a sinking feeling and looked at /dev/sda1 which is the boot partition on the system I haven't killed yet and, sure enough, grub-install had run on that drive. $ ls -lt /boot/grub total 2372 drwxr-xr-x 2 root root 12288 Nov 29 11:26 i386-pc drwxr-xr-x 2 root root4096 Nov 29 11:26 locale -r--r--r-- 1 root root7276 Sep 3 05:43 grub.cfg -rw-r--r-- 1 root root 2396122 Sep 3 05:43 unicode.pf2 -rw-r--r-- 1 root root1024 Jun 29 2019 grubenv drwxr-xr-x 2 root root4096 Jun 29 2019 fonts It didn't even touch any part of /dev/sdd1. With trepedation, I rebooted the good system and thankfully, it came right up since I hadn't modified /etc/default/grub. That was a bit of good luck but I thought it was supposed to write to /dev/sdd1 which would translate to /dev/sda1 when the drive was connected to the controller of the system that is presently belly up. I suspect the problem is the issue with the modules which another poster described. What am I failing to do to make the changes occur on the designated drive? Having it write this kind of stuff to drives other than the desired target is scary. Thanks for a good explanation and I may not be so lucky next time. Martin
Re: 780 files in /usr/share/zoneinfo/
Andy Smith writes:
> Hi Martin,
>
> On Sat, Nov 21, 2020 at 08:48:51PM -0600, Martin McCormick wrote:
> > find . -name "*" -exec ls -l {} \; \
> > |grep -F / \
> > | awk ' { total += $5 } END { print total }'
> >
> > That usually just adds the sizes of all the files it can
> > find all the way through the tree.
> >
> > If this is not an accurate way to determine how many
> > bytes there are in a directory then that would be the reason for
> > the discrepancy.
>
> The same file can be reached by multiple names. So by doing this you
> end up, in this case, with a ~256x amplification.
>
> A simple "du -sh" does a better job here!
>
> > cron only works in the time zone for wherever the TZ for the
> > system is set.
>
> Ah, I see. I've never tried it but I believe that systemd timers can
> have a time spec that includes time zone, so you can set timers that
> fire on a different time zone to that used by the rest of the system.
>
> $ systemd-analyze calendar '11:00 Europe/London'
> Original form: 11:00 Europe/London
> Normalized form: *-*-* 11:00:00 Europe/London
> Next elapse: Sun 2020-11-22 11:00:00 UTC
>From now: 6h left
>
> Cheers,
I do appreciate being corrected, here. What we really want to
know, here, is how much precious disk space is occupied by
whatever data base we are interested in. Links make copies of
files appear to exist when the data were only written once even
Martin
Re: 780 files in /usr/share/zoneinfo/
Michael Stone writes: > The kernel, compressed, is larger than that. The initrd needed to boot the > kernel is also typically larger than that. A modern system has more CPU > cache than that. At some point trying to save bytes is a waste of > developer > and administrator effort, and 3.5MB in 2020 is well past that point. If > you > want a minimal system, debian isn't for you. Instead, you'll need to hand > craft every file to make sure it isn't "wasting space". If that's your > thing, great. But it's just not a focus for debian. About the only place one still needs to think this way is with embedded systems where the computer is there to manage a machine of some kind, anything from a lathe to a food processor to a cement mixer, whatever . General-purpose computers are optimized to have as many resources as one can cram in to a higher and higher-density box so a few MB here or there aren't noticed but embedded systems are optimized with different priorities and one may discover that this box may be lightning fast but a bit skimpy on data storage. I am thinking of things such as cable TV boxes and dedicated audio-visual appliances that use DSP to emulate complex and expensive hardware by using mathematical algorithms that cause the system to decode digital TV signals or route internet traffic rapidly. If you aren't in to trying to modify some sort of embedded system to do something it wasn't originally designed to do then ram and storage are getting cheaper by the day and some things just aren't worth worrying about. Martin McCormick
Re: Fixing a Grub Foul-up, Not There Yet.
I am going to respond to one of my earlier posts as it doesn't help things at all to spread misinformation which I am guilty of, here. "Martin McCormick" writes: > I appear to be using grub, not grub2. No. It's grub2. Old Grub is now grub-legacy and is probably a dead fly on somebody's wall by now. > There's an extra little wrinkle in that, as a computer > warier who happens to be blind, grub isn't really accessible in > that you can not get direct access to it via a serial port and > terminal, say, from another computer that is actually working. I was simply wrong there. Of course, one has to have the lines now in /etc/default/grub that used to be in menu.lst but the idea is the same. If one has a serial port which is available at boot time, grub should work headlessly. The issue with the instance of grub that seems to have a run-time error due to a problem with modules, may be unusual because it may actually be a goner before it runs since it doesn't get to load it's modules. Martin
Re: Fixing a Grub Foul-up Solved!
David writes: > Your lack of success is because the the command you used has designed > behaviour to install the grub bootloader to the boot sector of > /dev/sdd, and also install the grub files you listed into the current > system /boot/grub (which was not on sdd at the time). That is the > reason why you see those files on /dev/sda1, because it was the boot > partition at the time you ran the command. Aha!! that makes perfect sense now that I think a bit. > > If you want a grub-install command that writes /boot/grub files > somewhere onto /dev/sdd then you will first have to mount the desired > target boot partiton of /dev/sdd on some mountpoint that you choose, > and then run a command something like this: > sudo grub-install > --boot-directory=/some/mountpoint/where/is/the/sdd/boot /dev/sdd I tried it and by Joe, it wrote to the correct drive so I don't feel quite so paranoid about using grub. Interestingly, I goofed again and grub ended up in what is the / directory instead of /boot/grub so it still didn't work but I knew immediately that the new grub was working because there was no error message about symbols not found, etc. I mounted the drive back on the good system and simply removed the grub from the top of the tree and then did the grub-install command again but thihs time, I installed to /mnt/boot. It wrote to /boot/grub as it should have and the system came right up. There was a minor glitch when I powered the box up as the keyboard reported a stuck condition and advised me to press F1 to continue. I thought this was a bit amusing since the keyboard appeared to be malfunctioning but F1 did the trick and about a tenth of a second later, I heard the beep that grub plays when the kernel is about to fire off. In short, the faulty grub that made it onto the system was all that was really wrong. Thanks to a lot of you, I appreciate grub more as this has been quite a little mini course in what it does. Martin
Cron Jobs and Time Zones Has Anything Changed?
In a recent discussion, someone indicated that there might be a way to set individual parts such as accounts on a unix system so that cron could use another time zone if needed to kickoff jobs on that system based on the time in another country. As far as I understand cron, one can set the system's time zone to only one value which is usually one's local clock time and that works very well since system logs and cron jobs all agree with what is appropriate for one's location. I record a news broadcast from one of the BBC services every week day at 17:45 British time. When Europe and North America stop or start shifting daylight in Autumn or Spring, there's a really good chance of missing some of the broadcasts if one doesn't think about it since these shifts don't all happen on the same time. One can certainly get the time anywhere as recently discussed by setting the TZ environment variable but, if you tell cron to trigger a job at 17:45, it only knows when that is based on the entire system's local time. Has anything changed recently to make this logic obsolete? In my case, I just have an old Linux box for which I set it's system time zone to Europe/London and call it good but this could get out of hand if one had more than 2 or 3 such schedules. One could also setup VM's if you have the memory to spare but this adds a lot of resource usage and complexity to the job at hand so my question is basically, has anything fundamentally changed in the way cron is used? This is not a complaint at all. I was first introduced to unix-like systems in 1989 and immediately knew that this was the sort of OS I wanted to stick with in amateur radio and technical tinkering in general. Martin McCormick WB5AGZ
Re: Cron Jobs and Time Zones Has Anything Changed?
Greg Wooledge writes: > I was vaguely thinking of a similar approach. Set up a job that runs > every hour, or across a set of hours that will cover all the possible > cases that you care about, in your crontab. Within the job itself, > set a TZ variable and determine the time in that time zone by whatever > means necessary, and then either abort or continue based on that time. > > This is also similar to how one approaches a complex date-based task > such as "run on the second-to-last day of each month". For that one, > you can set up the job to run on the 27th through the 30th every > month. Then within the job itself, determine whether it's the > correct day, and abort if it's not. What I was thinking of is a modification to cron which should integrate nicely with what cron already does. Crontab would have a new field at the beginning of each line which could normally be left at "default" which would be the normal behavior that we are used to. If someone wanted to run a job at 15:00 each day in let's say Japanese time, the first field of the line would read Asia/Tokyo instead of default. The line might look like Asia/Tokyo 28 15 * * 1-5 sh -c ". $HOME/.master.env;beep -f700 -l500" Since all unix-like systems start out with UTC and use those /usr/share/zoneinfo data base files to calculate what local time is, that information is handy in the time() function. localtime(time) gets you the adjusted number of seconds for your locale arranged in a structure containing fields for the current year, day of month, day of the week plus the hours, minutes and seconds with the adjustment for Summer or Winter for that zone. The sample line above would cause cron to grab the current epoch in seconds (UTC), feed that to localtimebased on rules for /usr/share/zoneinfo/Asia/Tokyo and then all those values to match the fields on the rest of the line. The time of 15:28 or 3:28 PM in Tokyo occurs around 01:28 in the Central US time zone and that's when your computer would have beeped. This would happen Monday through Friday Japanese time which still would be Monday through Friday in the US but early in the morning. If the time referenced in Tokyo had been 7:00 in the morning , it would happen the previous evening in the US. As long as one designated the time zone correctly and doesn't forget that 07:00 on Monday morning in Japan equals 5 or 6 o/clock on the evening of the previous day, it should just work. I haven't looked at the C code for cron, but I have written a few perl scripts that do things with time and dates and the current epoch-based number of seconds since utc Midnight January 1, 1970 is based on the C modules such that one's current wall-clock time is time(localtime). Just a thought Martin
Re: Cron Jobs and Time Zones Has Anything Changed?
> On Wed, Dec 02, 2020 at 01:58:45PM +, James B wrote: > > This might be wrong, but as far as I understand doesn't systemd > > now have the ability to manage cron jobs (as well as mount points, > > home folders and other things)?. Is there anything in this newer > > functionality that might make such a thing (re the request at the > > beginning of this thread) possible? > > Yes, I pointed this out to OP last time OP asked this exact question just a few days ago, so I don't know why they are asking again. I am guilty as charged but haven't yet found the relevant information as to how systemd helps solve this issue. Systemd is praised to the rooftops by some and cursed just as vehemently by others and it is what we are now using but it feels mostly like unix has always felt. I am obviously suffering from the worst sort of ignorance syndrome which can really bite in that sometimes, we have an idea what we don't know and other times, we don't even know that we don't know and that's really frightening. So, I need to read more general information about the differences between systemd and what we've been using up to recently. Martin McCormick
Re: Cron Jobs and Time Zones Has Anything Changed?
Andrei POPESCU writes: > On Jo, 03 dec 20, 07:39:14, Martin McCormick wrote: > > > > So, I need to read more general information about the > > differences between systemd and what we've been using up to > > recently. > > The Wikipedia page and/or https://systemd.io might be a good place to > start. > > Kind regards, > Andrei I've had a chance to investigate this more and the first thing I did was to go to wikipedia which told me about systemd, the on-going argument about it VS older ways to build unix-like systems and a time line when systemd began to take hold which was around 2015 when I retired from work so it kind of sneaked up on me and I didn't realize that I've been using it for 5 years, well, at least I didn't think much about systemd having different resources. I've used udev rules in making sound cards come up in the right order and that's totally a systemd thing. While searching for ways to use cron with different time zones, I found out about something called systemd timers just mentioned in this thread and it appears they definitely will do the job but one must magically have made the connection between this concept and the concept that cron, the usual go to resource for making lighting come on at a certain time, purging stale files or backing up the system now has a new kid on the block called systemd timers. One can even list all the active systemd timers https://www.maketecheasier.com/use-systemd-timers-as-cron-replacement/ $ systemctl list-timers And, sure enough, I had a number of them ticking away. An on-going problem about self-education is that it's easy to limit the scope so much that we miss connections. Systemd timers doesn't even sound like a replacement for cron but think of it as cron on steroids. One of the things in the wikipedia article about systemd was a complaint by someone that it's just too complicated. All I can say is that maybe or not that is true but that's just life. Start simple. Make more and more complexity as problems with simple show up. One day, try to make things simple again. Lot's of luck with that. I'm not so sure that being blind makes nearly as much difference these days as it once did, but making connections that relate one knowledge base to another in a meaningful way will always be a problem for us. Martin
Re: Web-bot tarpit aka spider trap (was: swamp rat bots Q)
It's nice to see that I am not the only sick puppy out there.
At 69 years old, I still don't have much trouble with
getting in touch with my inner 12-year-old when it comes to
intrusive marketing which is so prevalent these days.
I found our old dial-up modem in a box of odds and ends 2
years ago and wondered if it could read callerID tones sent after
the first ring. It can so I started on a perl program that
initializes the modem for callerID and then compares the strings
received with a pair of files, one of which is called scum and
contains callerID name packets of folks we don't want to talk to.
The other is called good and looks for names of friends or anyone
else we like hearing from. It is actually scanned first and
causes the program to abort.
If anyone's name matches a name in scum, or the caller's
ID appears to be blocked or, in one case, matches a whole
exchange, (first 3 digits after the area code), I call a
subroutine that makes the modem answer for half a second then
drops the call.
We were bombarded with garbage calls all day long until
the US presidential election and it was so satisfying to hear the
program kill the call by answering just as the second ring began.
I also installed subroutines that looked for the word
"SPAM?" just before the name or "ROBO?" also just before the
name.
We now get very few unwanted calls but occasionally,
we'll get a call from lala-land from someone we don't know who
lets the phone ring until the answering machine picks up and then
fails to leave a message and I tell my wife, "I'll go put them on
the juke box." especially if it looks like the name of a business
with which we have no relationship.
Our phone is very quiet these days except for legitimate
calls.
Martin McCormick
Re: Gene you poor soul
Gene Heskett writes:
> On Tuesday 21 May 2019 08:07:17 pm Jude DaShiell wrote:
>
> > If brltty is killing all other usb numbers, that's a bug that needs
> > fixing! Much of the other computer peripherals going into modern
> > computers these days go in through usb ports.
> > If that ran, my standard keyboard and speakers would be killed and if
> > those two get killed, I can't use a computer since I use speech
> > synthesis for output since I've never been able to see a screen.
> >
> This only affected the seriel to usb dongles. And I've at least 3 in
> this system. Since x10 will never make another CM-11a, and a
> firecracker is different enough it won't talk to any of the older stuff,
> it either works or I build a whole new system to automate one group of
> lights.
>
> > One time I remember the Navy went and fixed band width hogging bugs
> > with their mandatory training and took out the requirement to position
> > a mouse and click to move to the next slide. This was done to support
> > accessibility. Its other effect was on all of the ships in the
> > fleets. Sailors suddenly discovered it was easier to complete their
> > mandatory training.
> > That was one instance where accessibility improvement had a good
> > effect on those in remote environments.
> >
I found this thread accidentally and got interested in it
so here are my two Cents worth.
When you run a program or application, that application
may need kernel modules that aren't part of the kernel right now
so it looks in a special directory which is
/lib/modules/long_number. In my case, long_number is
4.9.0-6-686-pae
Yours could be the same or more likely, it will be
different and you find that out with the command
uname -r
The output you get from uname -r is what you put after
/lib/modules/
Your modules directory might have a module supporting
brltty and you can find that out by cd
/lib/modules/that_long_number and then doing ls |grep -i brltty.
What that does is list all the files in the directory and
only print those containing brltty in their name and it won't
matter if the brltty is all upper case, lower case or some crazy
mixture.
If nothing is there, you have no brltty modules so you
have found the bottom of that rabbit hole and don't need to dig
any further.
If you do, however, then lookup how to keep your computer
from using that module since you don't need brltty anyway. That
procedure is different with different versions of Linux so read
how your version blacklists kernel modules to keep them from ever
loading.
This is one of the many things about unix that are really
useful since you can tweak your system to fix seemingly
intractable problems at times.
Martin McCormick
Upgrade to Buster and perl Device::SerialPort
There is a system here that I upgraded to buster from stretch which runs a perl program I wrote that accesses a RS-232 dialup modem over /dev/ttyS0, a native RS-232 port on the mother board. It works. The dialup modem is pretty useless these days for it's original intended purpose but I wrote a nasty little program that reads callerID tags and lets telemarketers take a trip down Memory Lane when they repeatedly call us. There is another system whose hard drive crashed a couple of weeks ago so I put stretch on it from a netinstall disk and promptly upgraded it to buster before installing cpanplus and Device::SerialPort as the less software you have, the smoother the upgrade, generally. Interestingly enough, the system that didn't crash and was upgraded perl and all from stretch to buster is handling the modem fine, whistling cheerfully at callers who are too big of a wuss to properly identify themselves. On the system that got perl after the new drive, the complaint is that none of the parameters such as baud rate, handshake and all other trates throw errors right and left about uninitiallised variables, the variable being the port that was previously defined. It's like the port isn't there. The first program I tried that exhibited this behavior was one that accesses /dev/ttyACM0, a device that shows up if you plug a Uniden scanner radio in to a usb port. In stretch days, it nicely acted as a RS-232 port. Right now, it shows up as soon as the cable is connected but the program I wrote in perl to control it throws the same errors. Obviously, the two boxes are separate hardware and the cable to the radio scanner is not long enough to stretch over to that same box to see if it works there, but I suspect it might. I think the difference is that perl was added after the buster upgrade on the serially-deaf box and neither the virtual usb serial port nor the actual hardware serial port seem to work any more even though they show up in /dev and the /dev/ACM0 port goes away when unplugged. Any ideas as to what is going on here? I am also a member of a perl discussion list but I thought I would start here as this problem is directly related to buster VS stretch. With stretch, it all worked. The only other possibility is that there may be a recommended syntax for perl and serial ports that I got by with violating in earlier Linux and now the chickens have come home to roost and I need to do it differently but much of what I wrote is stolen from other code examples so I bet it is something in buster that isn't right yet. Any suggestions are much appreciated. Martin McCormick WB5AGZ
Re: Upgrade to Buster and perl Device::SerialPort
David Christensen writes:
> Post a console session showing the commands issued and the error messages
> produced.
I'll go you one better. I'll show the code of one of the
programs that dies instantly. It uses /dev/ttyACM0 which is the
device created when a Uniden scanner radio is plugged in to a usb
port. I also have the program that controls a RS-232 modem and
uses /dev/ttyS0 and gives the same results but contains more
lines so let's go with the scanner program.
Here's the console session short but not sweet.
Script started on 2019-07-10 12:32:35-05:00 [TERM="linux" TTY="/dev/pts/7"
COLUMNS="80" LINES="25"]
1wb5agz martin tmp $ p2 sts
Can't call method "baudrate" on an undefined value at /home/martin/etc/p2 line
37.
2wb5agz martin tmp $ ls -l /dev/ttyACM0
crw-rw 1 root dialout 166, 0 Jul 10 08:32 /dev/ttyACM0
3wb5agz martin tmp $ exit
Script done on 2019-07-10 12:33:21-05:00 [COMMAND_EXIT_CODE="0"]
You used to only see that kind of response if /dev/ttyXXX
was missing. On rare occasions, the scanner radio's tty would go
away and one would need to umplug it and reseat the plug to get
it back and it would then work.
The rest is 66 lines of code for the scanner control
program.
#!/usr/bin/perl -w
use strict;
use warnings::unused;
use File::Basename;
use File::Copy;
use File::Spec;
use Time::Local;
use Device::SerialPort;
sub waitfor {#Receive from serial port.
my $received = "";
our $port;
until ( "" ne $received ) {
$received = $port->lookfor;
}
return ($received);
}#Receive from serial port.
#no more sub routine code past this point
#main global variables
our @dbstrings;
#main local variables
my $response;
my $cmd= "";
my $ONEARG = "";
#my $counter = 1;
#Setup the comm port.
my $dev = "/dev/ttyACM0";
our $instring = "";
our $port = Device::SerialPort->new("$dev");
$port->baudrate(115200);
$port->databits(8);
$port->parity("none");
$port->stopbits(1);
$port->handshake("none");
$port->read_const_time(500);# 500 milliseconds = 0.5 seconds
$port->are_match( "\r", "\n", );# possible end strings
foreach $ONEARG (@ARGV) { #each argument
$ONEARG =~ s/^\s+//;
$ONEARG =~ s/\s+$//;
$cmd = $cmd . " " . $ONEARG;
$cmd =~ s/^\s+//;
$cmd =~ s/\s+$//;
}#each argument
$cmd = uc $cmd;
$port->write("$cmd\r");
$instring = waitfor;
my @chars = split( "", $instring );
foreach my $char (@chars) {#
#print ord($char);
if ( $char =~ /[\,\s\.\-\+\_0-9A-Za-z]/ ) {#printable
$response = $response . $char;
}#printable
}#
print "$response\n";
$instring = "";
exit(0);
solved! Upgrade to Buster and perl Device::SerialPort
After making an utter ass of myself on this group and
another list, something somebody suggested helped me solve the
problem and the news is good for all.
I thought it was odd that a serial program I am running
on one buster system was running like the proverbial house on
fire, no trouble at all while that same code blew sky high on
another buster system that I had had to replace the boot drive on
after it crashed.
We need to be careful when diagnosing trouble not to
think about esoteric issues and be sure the power plug is in the
outlet and the ON/OFF switch is in the On position.
Well, it wasn't quite that stupid on my part but darn
close.
For as long as I can remember, debian Linux defaults to
disallowing non-root users in general from accessing the serial
ports. To access /dev/ttyXx, non-root users need to be added to
the group "dialout" and an upgrade doesn't remove anybody who is
there now from dialout so one doesn't even have to give it a
second thought.
In this case, this was more like starting from bare metal
and I simply forgot to add me to dialout which is done by
sudo usermod -a -G dialout martin
One must then log out and back in for the change to take
effect.
A person on a perl discussion group reminded me that I
should always test for issues like this which is absolutely
correct. Perl will tell you about the error and kill the program
as it should but you need to be smart enough to put the test in
or you get what I got which is utter confusion because the
program blows right through the bash: /dev/ttyS1: Permission denied
condition and tries to grind on which is impossible since the
port won't work. There is no reason to continue after that
point.
The function in perl is called die and it has the ability
to print a string stating what error caused the death but it
doesn't just appear by magic.
I added the dialout group and that program now also runs
like the blazing house. Nothing's wrong with buster and serial
ports after all.
I use the die("cause of death\n") function all the time
but forgot to put it in to a test to see if one could access
/dev/ttyXx and it bit me.
Thanks to all who responded with good suggestions.
Martin McCormick WB5AGZ
Looking for a Couple of Programs in Buster
I wanted ckermit and a program called dosemu and apt-get says they aren't available. Here is the sources.list file I am using. Are these applications really gone or is there another archive to search sources.list follows and I started out in stretch: # # deb cdrom:[Debian GNU/Linux 9.9.0 _Stretch_ - Official i386 NETINST 20190427-11:18]/ buster main #deb cdrom:[Debian GNU/Linux 9.9.0 _Stretch_ - Official i386 NETINST 20190427-11:18]/ buster main deb http://ftp.us.debian.org/debian/ buster main deb-src http://ftp.us.debian.org/debian/ buster main deb http://security.debian.org/debian-security buster/updates main deb-src http://security.debian.org/debian-security buster/updates main # buster-updates, previously known as 'volatile' deb http://ftp.us.debian.org/debian/ buster-updates main deb-src http://ftp.us.debian.org/debian/ buster-updates main Thanks. Martin McCormick WB5AGZ
Re: Looking for a Couple of Programs in Buster
Greg Wooledge writes: > https://tracker.debian.org/pkg/ckermit > https://tracker.debian.org/pkg/dosemu > A very useful site, in my experience, and tragically not well known. That is really good information. Thanks. I am not even sure if dosemu will do what I need and ckermit was always my goto terminal when I need to blast bytes at a radio scanner or anything else using serial comm. Anyway, as a computer user who happens to be blind, I have a DOS diskette that reportedly has the ability to make your PC boot off of a usb port even if such capabilities are unknown to the BIOS. Maybe, and I stress maybe, dosemu might let one hear the menu to know which key to press. When one actually uses the disk, one isn't going to hear that menu but you just wait for the disk to stop turning and press the desired key. Being able to hear the menu would depend on whether or not the emulator has ASCII characters to work with or just bit-mapped pixels. The ultimate goal here is to be able to usb-boot a thumb drive containing debian wheezy as I have some development tools that compile in that wheezy version and do not compile in today's world. When I am not using the PIC development system, I could boot normally in to buster or stretch on that same system. I don't like to throw out older PC's as long as they can still do useful work. This particular system is a 600-MHZ Pentium with a GB of RAM, too slow to do gnome but good enough for email, audio and general tinkering. Again, thanks for the tracker info. Martin McCormick
Can Grub Boot Through a USB Port on an Old PC?
The PC is old enough that it can not natively boot via a serial port but it might if grub knew about the bootable drive sitting in one of the ports. Thank you. Martin McCormick
Re: Can Grub Boot Through a USB Port on an Old PC?
Pascal Hambourg writes: > Le 23/07/2019 à 04:53, Martin McCormick a écrit : > Do you mean that GRUB is installed on an internal drive ? Yes. > By default, GRUB relies on the BIOS disk services to access drives. But it > also has native ATA and USB drivers which are not loaded by default. See > the 'nativedisk' command. > > > > Be aware that loading native drivers switches GRUB to a new different > world > : BIOS drives (hdX) will not be available any more and environment > variables pointing at drives such as $root and $prefix will need to be > updated with native drive names (ataX, usbX...). Thanks again for the heads up. > > Also, note that chainloading a bootloader (boot sector or ntldr) which > relies on BIOS disk services will not work. Loading a kernel and initramfs > should work if they include native USB drivers. My thanks to all who responded. In addition to your responses, I later found a discussion on one of the stackexchange sites dealing with this same issue and was about to kick myself for wasting everybody's time but it turned out every response touched on a different aspect of the issue so it sounds like it will keep me busy for awhile. It may turn out to be less of a headache to make it a duel-boot system. One boot would be the latest debian console and the other would be Debian Wheezy as I have some PIC microcontroller development tools that don't make in today's world. This is a Dell Dimension motherboard with 1 GB of RAM and a 600-MHZ Pentium which works very well in the command-line world so I'll run it until the motherboard fries. Yup! It's about 20 years old but no reason to trash it yet. Martin WB5AGZ
Re: Can Grub Boot Through a USB Port on an Old PC?
Pascal Hambourg writes: > If the system on the USB drive is GNU/Linux, you just need to put /boot on > the internal drive so that GRUB can load them. > > > > as I have some PIC > microcontroller development tools that don't make in today's > world. > > > Did you consider chroot instead of multiboot ? Not until I read this message! I have used chroot a few times and know that the directory is now the root of the jail you have made but I didn't stop to think that this also means that while one is in that jail, all the pathsare now relative to the bounds of the jail including libraries and executables which also see only wheezy libraries so any executables should be happy. The binaries we make with the assembler never execute on site but do so in the tiny brains of the PIC processors. The programming device we use to flash the PIC eprom connects to a native serial port or a usb-to-serial converter and if those work through the walls of the jail, there should be no trouble. If the programmer still works, that is a winner and many thanks. Martin McCormick
Re: Can Grub Boot Through a USB Port on an Old PC?
Pascal Hambourg writes:
> Did you consider chroot instead of multiboot ?
I have had some time to play with chroot and there is only 1
problem I have just discovered with everything else working
perfectly.
What I did was use what would have been the wheezy boot
drive as follows:
The only partition that could be used here was partition
1 which, on this system was /dev/sde1.
I mounted /dev/sde1 to /mnt and used rsync to copy
everything it could including devices to /home/wheezy
rsync --devices -alHvbq /mnt/ /home/wheezy
It appears to have done that and /dev/ttyUSB0 was there
so I tried to test it as user martin, not root as in "Safety,
first."
When I tried to access the port, I got Permission denied
which means one isn't in the dialout group. This always happens
if you forget to add yourself via usermod as root.
I went backup to root while still in jail and commanded:
usermod -a -Gdialout martin
There was no complaint so I su'd back to martin and did groups martin.
I was in all the groups one is put in by default when
creating a new account but no dialout.
After trying a full logout from the jail, I did what
many criminals do and went right back in and tried groups martin
once again.
No dialout and no complaint either.
I did successfully compile a small PIC binary from source
with the old environment and compared it with a compile from 2016
and noticed that the binaries were different. After saying a few
choice words, I checked a little further and noticed that both
were exactly the same size to the byte so I ran strings on each
binary and found or maybe rediscovered that gpasm which is the
PIC assembler puts a time and date stamp on each build you make
from source so I don't know for sure that the new build is
exactly the same but I would bet money it is.
Anyway, usermod from jail fails quietly and I never get
added to dialout.
Again thanks for the suggestion and any constructive
ideas are appreciated.
Many thanks.
Martin McCormick
Re: Can Grub Boot Through a USB Port on an Old PC?
Tixy writes: > I've used 'schroot' in the past for this sort of thing, let's you > configure what to mount and I believe either defaults or has examples > for the common things you're likely to need like /dev and /proc. I > don't remember the details as it's been quite some years since I used > it. Thanks for this information. I have installed schroot as this seems to be the missing link I didn't know about. I knew one had to have all the resources turned on so to speak and it looks like schroot makes getting it right in this century much more likely. martin
Easiest way to do VGA to Text
I've been trying to do the impossible, more like the impractical, for some time now so I need a knowledge infusion. I want to be able to read the VGA output of a computer, do OCR on it and have ASCII text. As a computer user who happens to be blind, one of the most frustrating issues is the fact that except for expensive servers, none of these boxes output any machine readable text when booting up or in setup mode such as when the coin cell that powers the CMOS BIOS gives up the ghost and one needs to do a setup on it, etc. It seems as though we may have reached one milestone in that one can buy a usb frame grabber that spits out UVC webcam-style video. The representative of the company which makes the device told me that most common flavors of VGA cards produce signals that would work with the device but some odd-ball cards won't work with it which makes sense. Assuming most will work, what one would have is frame after frame of digitized video. If this is a setup screen, the only thing likely to be changing until you do something is the cursor may be blinking otherwise, it's going to be pretty stable. You'd have one frame of video to do the OCR on and then one does something such as hit the Tab or one of the arrow keys and then you grab another frame and read it and so forth. Are there any free projects out there which take the raw video as input and output text as output since the frame grabber is just the beginning of the beast and then you have to convert it to text and maybe some method of determining where the highlight as in cursor position is so as to know what one is about to select? Needless to say, but saying it anyway to avoid confusion, one would have the frame grabber and text engine on a different working debian computer since the sick one isn't capable of doing much until the BIOS gets set correctly. I have 4 older PC's that generally work well running debian but Right now, 3 of them need varying degrees of attention to their BIOS setups as Dell motherboards and possibly other brands will occasionally modify their boot sequences for some reason and the only way one can boot from a CDROM is to get in to the BIOS setup and yank the boot order back to one where the CD drive is ahead of the hard drive or put an unbootable hard drive in. Six or eight months later, one will suddenly discover that the boot sequence has fallen back to the useless one where the floppy drive is first, followed by the hard drive followed by the CDROM. Martin McCormick WB5AGZ
Re: Wireless home LAN - WiFi vs Bluetooth?
ng power supplies that ran this data network and each supply was attempting to still run. The chopping oscillators in those supplies normally chopped at ultrasonic frequencies in the range of 100 KHZ or more but that day, there were probably only receiving 20 or 30 volts at the power socket so they were oscillating at very audible frequencies like so many derange smoke alarms, rising and falling in pitch in an eerie wale. Our serial RS-232-based network seemed sophisticated at the time with battery backup of each node and so forth but that sustained period of low and varying voltage convinced the processors in the nodes that someone was changing baud rate and loopback settings so, when the phase was restored and the normal world returned, that network was fouled up beyond recognition for quite a long time. We got the nodes back by reloading each one with saved backup files so that wasn't so bad but some of the weirdness was propagated to nodes in other buildings so we didn't get it all cleaned up for literally weeks. It was mostly okay that afternoon but we would get calls from departments across the university that this or that printer or VT100 terminal wasn't working when somebody tried to use it and the fault turned out to be another randomly deconfigured port. The down phase created an apocalyptic scene what with overhead lighting mostly absent but occasionally trying to flicker on and the wailing power supplies. Martin McCormick WB5AGZ
Re: 3 phase power (was Re: Wireless home LAN - WiFi vs Bluetooth?
rhkra...@gmail.com writes: > An update | correction | recollection ;-) > > On Tuesday, July 30, 2019 11:34:43 AM rhkra...@gmail.com wrote: > > I have seen diagrams in NEC code books for a different arrangement to > get > > 120 volt 3 phase power, but I don't recall ever actually encountering > that > > in real life. > > Oh, wow, how quickly I forget -- I did encounter systems like that, often > for > lighting in industrial applications, And, further, iirc, we could (and > did) > buy and install florescent light (and maybe HID?) fixtures designed to > work on > 208 volts, which we connected phase to phase in that kind of system. > > Of course, I could be misremembering. > > > In that case, iirc, 120 volt loads are connected from a hot > > tap to the neutral wire (the 4th wire of the 3 phase arrangement), and > you > > get (nominally) 208 volts (RMS) connecting phase to phase. I have seen > > things like motors that were rated like 240 / 208 volts (or something > like > > that). > That's the way I understand it too. One of the buildings we worked in at Oklahoma State University had low-voltage wiring to the light switches. Pressing the On button turned on the overhead light, accompanied by the click of a latching relay. Pressing Off released the relay with another audible click from above. I asked one of the electricians as to why the fancy light switches. The building had 3-phase wiring for the air handlers and the lights were powered phase-to-phase at 208 volts. This would have been dangerous if someone took the plate off of a switch for any reason because every wire would be hot between it and anything else. Martin
Re: Easiest way to do VGA to Text
john doe writes: > On 7/30/2019 7:01 PM, Martin McCormick wrote: > > > > I have 4 older PC's that generally work well running > > debian but Right now, 3 of them need varying degrees of attention > > to their BIOS setups as Dell motherboards and possibly other > > brands will occasionally modify their boot sequences for some > > reason and the only way one can boot from a CDROM is to get in to > > the BIOS setup and yank the boot order back to one where the CD > > drive is ahead of the hard drive or put an unbootable hard drive > > in. Six or eight months later, one will suddenly discover that > > the boot sequence has fallen back to the useless one where the > > floppy drive is first, followed by the hard drive followed by the > > CDROM. > > > > Why not "simply" deconnecting the hdd once booted reconnecting it? > Granted, it's clearly not ideal! :) I chuckled when I read your message. Not because there was anything at all wrong with the suggestion but because that was the first thing I tried years ago when I stumbled across the problem. Since the boot drive was disconnected, the computer just gave me that nasty little double high-pitched beep which says the game is over before it even started. It seems that there has to be a drive in place that would boot if only it could. Again many thanks. Martin
Re: Easiest way to do VGA to Text
David Christensen writes: > What about using a computer whose CMOS Setup utility is accessible via the > serial port? This article indicates the Dell 2450 is capable: > > > http://tldp.org/HOWTO/Remote-Serial-Console-HOWTO/rhl-biosserial.html > > > David > > Thanks for the information but None of the Dells I have are that model. They are all traditional eyeballs-only style. I even have a card that was manufactured many years ago that replaces the VGA card in a number of PC models but not, of course, the ones I have. I had bought it to use at work and it did work fine in a system there but the card will not physically fit the connector on the one system that has a removable video output card and the others all have VGA integrated in to the mother board. You fight the war withthe weapons you have, not what you wish you had. The other possibility touches on a comment that another poster to this thread mentioned. I do own an iPad which has a commercial OCR program on it that while driving one crazy with it's limitations, can sometimes work quite well. If you use the iPad's rear-facing camera and shoot a photo of the monitor screen, it sometimes produces useful results but it is normally kicking and screaming every step of the way. The OCR program is really meant to photograph pieces of paper under light that doesn't flicker. The monitor screen is a classic CRT which flickers at the frame rate of 60 FPS with no setting to increase exposure time which might solve the problem. I waited until night, clamped the iPad in to a tripod fixture aimed at the CRT and observed that it continuously cycled in and out of error mode due to the flicker. A LCD-type monitor might work much better as they have almost no flicker. This is definitely one of those projects in which one must not let perfect be the enemy of good enough. Thanks for the good ideas. Martin
Re: Easiest way to do VGA to Text
Felix Miata writes: > Are these old Dells continuously connected to power whether booted or not? Yes. > I have a bunch of old Dells. When left unconnected to power, some wear > down the > 2032 coin cell CMOS batteries with unusual haste. Once this happens and > the > battery is replaced, reconfiguring the BIOS is required, unless 100% use > of > defaults proves acceptable, which because of the date and time at least, > for me is > never. I know what you mean. Some time before I retired in 2015, we had a Dell server at work that must have had bad steering diodes somewhere on the mother board because it could drain one of those 2032's in just a few hours. Nobody even knew there was a problem until it was shut down for some length of time and then it would come up wrong. Trying to remember what all happened is a bit difficult now but I seem to remember that one could power it up and it would start immediately for 10 or 15 seconds and then shut off at which time one could press the Start button and it would run with all the defaults. Good servers just remember what power states they were in when last shut down or maybe there is a BIOS setting that you can set to determine this behavior but there is none of this business of running for a few seconds and then going in to shutdown. The department that actually owned the server decided to replace it with a new one and that's the end of that story as far as I can relate. The university would have put the defective server in Surplus and state law required that it be sold at auction at which point, some poor bloak would buy it and make the same discovery we had made.:-) Martin
