hard disk device name change after reboot
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 hi list, I run Debian 4.1.2-13 on Dell PowerEdge 650 with kernel 2.6.21-2-686. When I install the debian, the device name of partition is hda1, hda2. After install, I reboot it for times and it works. One day, I find it fail to reboot, after waiting for root filesystem for a while, Debian drop to shell(busybox). Then I notice the device name of hard disk has automaticly changed to hde1,hde2. some times, the devices name is changed to others like hdg1, hdg2, etc. after reboot. What's the problem then? Can I fix the device name? Any clue for it is appreciated. :-) -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.7 (MingW32) iD8DBQFHLdwNJo9Njjkvy34RAmL5AKDopvZOV7jtDK/U83ub65cRDoh+QwCfSKOy kklg+7i10HxX6OLvGa2gtHo= =+eDU -END PGP SIGNATURE- -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: hard disk device name change after reboot
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 It seems it works with some filesystem and do *NOT* work for swap partition. here as I show below, hde1 is swap partition and there are no label or uuid exists. as I know, tune2fs can add uuid or label for ext2/ext3 partitions. How can I add uuid or label for swap partition then? Another question is, can I add uuid as device name in grub's menu.lst? (something like kernel /boot/vmlinuz-2.6.17-2-686 root=/dev/sde2 ro) cc:~# /lib/udev/vol_id /dev/hde1 ID_FS_USAGE=other ID_FS_TYPE=swap ID_FS_VERSION=2 ID_FS_UUID= ID_FS_UUID_ENC= ID_FS_LABEL= ID_FS_LABEL_ENC= ID_FS_LABEL_SAFE= cc:~# /lib/udev/vol_id /dev/hde2 ID_FS_USAGE=filesystem ID_FS_TYPE=ext3 ID_FS_VERSION=1.0 ID_FS_UUID=b9599834-8d37-41fb-9389-c3394c9ea21f ID_FS_UUID_ENC=b9599834-8d37-41fb-9389-c3394c9ea21f ID_FS_LABEL= ID_FS_LABEL_ENC= ID_FS_LABEL_SAFE= cc:~# tune2fs -l /dev/hde2 | grep UUID Filesystem UUID: b9599834-8d37-41fb-9389-c3394c9ea21f cc:~# Owen Townend 写道: > On Sun, 2007-11-04 at 13:55 -0800, Andrew Sackville-West wrote: >> On Sun, Nov 04, 2007 at 10:49:49PM +0800, hhding.gnu wrote: >>> hi list, >>> >>> I run Debian 4.1.2-13 on Dell PowerEdge 650 with kernel >>> 2.6.21-2-686. When I install the debian, the device name of >>> partition is hda1, hda2. After install, I reboot it for times >>> and it works. One day, I find it fail to reboot, after waiting >>> for root filesystem for a while, Debian drop to shell(busybox). >>> Then I notice the device name of hard disk has automaticly >>> changed to hde1,hde2. some times, the devices name is changed >>> to others like hdg1, hdg2, etc. after reboot. >>> >>> What's the problem then? Can I fix the device name? Any clue >>> for it is appreciated. :-) >>> >> Not sure specifically what is causing this problem, but in the >> meantime, switch to using LABEL's or UUID's to identify disks. >> That way you can continue booting until you figure it out. >> >> A > > Hey, To elaborate, vol_id can get the uuid for a volume which you > can then use in fstab. Eg this ntfs sata disk partition I called > 'Y' in windows. $ sudo vol_id /dev/sda1 ID_FS_USAGE=filesystem > ID_FS_TYPE=ntfs ID_FS_VERSION=3.1 ID_FS_UUID=4862AAF824AADB2F > <--This part ID_FS_UUID_ENC=4862AAF824AADB2F ID_FS_LABEL=Y <--Or > this part ID_FS_LABEL_ENC=Y ID_FS_LABEL_SAFE=Y > > This then translates to a line like this in /etc/fstab > > # /dev/sda1 UUID=4862AAF824AADB2F /ntfs/Y ntfs-3g > rw,user,users,gid=users,umask=0002,en_AU.utf8 0 0 > > or 'LABEL=Y' would also work. > > Doing this for all volumes renders device name changes a non-issue. > > > Hope this helps. cheers, Owen. > > PS/NB: My linux partition uuids look slightly different e.g. > e029093d-b95f-434a-b005-ab4f807e2ca8 As generated by uuidgen. > > > -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.7 (MingW32) iD8DBQFHLn9sJo9Njjkvy34RAlgdAJ9dgWsVNFqNKcAA8DDgBFN/sQPizQCfdwNn Cxd1BCC8Se7pdPlPmvMKQsk= =0Vv3 -END PGP SIGNATURE- -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: hard disk device name change after reboot
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 I try to discover why the device name changed and find something. I run lspci and find 2 ide interface, as show below: 00:05.0 IDE interface: Silicon Image, Inc. PCI0680 Ultra ATA-133 Host Controller (rev 02) /sys/block/hda/device/driver->../../../../../bus/ide/drivers/ide-disk 00:0f.1 IDE interface: Broadcom CSB6 RAID/IDE Controller (rev a0) /sys/block/hda/device/driver->../../../../../bus/ide/drivers/ide-cdrom So, I disable the cdrom in bois. The hard disk name is /dev/hda now. Question is: what is the order kernel detect the hardware? when will it detect cdrom before hard disk and assign name hda to cdrom? Andrew Sackville-West 写道: > On Sun, Nov 04, 2007 at 08:26:03PM -0800, Andrew Sackville-West > wrote: >> On Mon, Nov 05, 2007 at 10:26:52AM +0800, hhding.gnu wrote: >>> It seems it works with some filesystem and do *NOT* work for >>> swap partition. here as I show below, hde1 is swap partition >>> and there are no label or uuid exists. as I know, tune2fs can >>> add uuid or label for ext2/ext3 partitions. How can I add uuid >>> or label for swap partition then? >>> >> sorry no help on swap. >> >>> Another question is, can I add uuid as device name in grub's >>> menu.lst? (something like kernel /boot/vmlinuz-2.6.17-2-686 >>> root=/dev/sde2 ro) >> root=uuid=/dev/sde2, I think. > > ummm... that should be > > root=uuid=, I think. > > > A > >> A > > > -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.7 (MingW32) iD8DBQFHLshyJo9Njjkvy34RAvT7AKDmPwjDdf0iBoNvV9aBMfOkyh8J1QCgtbUK 0su2uBp2VBLezHGNM8WM0oc= =+Sm8 -END PGP SIGNATURE- -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
how can I aptitude install without interaction
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 When I upgrade libpam0g，a dialog is show by aptitude to make me select services to restart. I want upgrade it without interaction. I have tried aptitude -q=100 -y -f install libpam0g，but it does not work. -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.7 (MingW32) iD8DBQFHL85JJo9Njjkvy34RAm3qAJ4jB0psMvZn4h3qtfRvYOrdkX1g/ACeNREs F+D98nEolFWD7G7LKRu8S1Y= =EQBL -END PGP SIGNATURE- -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
nvidia package for xorg doesn't work
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Is it bug for debian package? Should it be nvidia_drv.so? I run m-a a-i nvidia , and the X fail to start. the error message is can't find driver [EMAIL PROTECTED]:~$ apt-file search nvidia_drv nvidia-glx: usr/lib/xorg/modules/drivers/nvidia_drv.o nvidia-glx-legacy: usr/lib/xorg/modules/drivers/nvidia_drv.o After I download and install nvidia driver from homepage of nvidia, the file name is nvidia_drv.so, It works for me. [EMAIL PROTECTED]:~$ locate nvidia_drv /usr/lib/xorg/modules/drivers/nvidia_drv.so [EMAIL PROTECTED]:~$ -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.7 (MingW32) iD8DBQFHMXrSJo9Njjkvy34RAqqeAKCE2JMz/p0imgUf1cABCxUV9ZsGBACfR8Ea kw5pVJR2yUf0GHeLB5YdFlA= =di7T -END PGP SIGNATURE- -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Can't start X-Server as normal user
[EMAIL PROTECTED] wrote: # startx ...works $ startx ...does not In /home/sam/.xsession-errors I can see Xsession: X session started for sam at Fre Jul 25 09:07:21 CEST 2008 open: Permission denied Now, I saw somewhere on the web[1] that chmod 1777 /tmp might solve the problem. I remember that I was not able to write to /tmp after the problem came. I will try this when I am back home after work and let you know. sam [1] linuxforen.de For directory /tmp, you should chmod 777 /tmp and chmod o+t /tmp[1] The t means stick, man chmod for detail. [1] [EMAIL PROTECTED]:~$ stat /tmp File: `/tmp' Size: 2048Blocks: 4 IO Block: 4096 directory Device: 808h/2056d Inode: 2 Links: 6 Access: (1777/drwxrwxrwt) Uid: (0/root) Gid: (0/root) Access: 2008-07-28 20:55:21.0 +0800 Modify: 2008-08-06 10:50:21.0 +0800 Change: 2008-08-06 10:50:21.0 +0800 [EMAIL PROTECTED]:~$ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Dual Core processor ?
Account for Debian group mail wrote: Hello, I'm in the process of replacing one of our mail servers with a new machine. Is Debian stable enough on the Intel or AMD dual core 64 bit processors for a mail server? This machine will get heavy use and runs hard for 24 hrs a day. Any suggestion for something fast but very stable? Thanks, Ken Yes, you can run etch without any problem. Most the server support debian etch as I tested. And, lenny will be stable in september. :-) -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
when debian run out of memory
Hi list, When debian run out of memory, I can only ping the host and can't ssh to the host. It seems oom-killer is running but memory is still exhaust. What should I do then? Only reboot can solve the problem? Can I protect oom-killer from kill the sshd then I can ssh to the box to kill the process? Ding Honghui -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: running CGI scripts in apache? please help
Zach Uram wrote: I noticed darcsweb put the CGI file in /usr/lib/cgi-bin/darcsweb.cgi /usr/share/doc/darcsweb/README.Debian says I should be able to access my darcs repo which i made in /var/ww/darcs/ by going to http://www.mydomain.org/darcsweb yet when I do nothing happens, so how do I get it to use the file in /usr/lib/cgi-bin? I noticed my /var/www/cgi-bin/ is empty. Runing Apache 1.3 on Debian 4.0 Zach Should /var/ww/darcs/ be /var/www/darcsweb/? /cgi-bin/ is alias to /usr/lib/cgi-bin, so /var/www/cgi-bin should not be there. Regards, Ding Honghui -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: How to find dependency information for a package ?
NewDeb wrote: See, this is a mystery to me, because I am running the commands and viewing the control file of one single package. Why is the version showing up separately ? Ok, so what I am actually after is this - I am going to get an input of several .deb files (of proprietary software within my company). I have to read the .deb files to extract the payload and bundle them together to form a "big package" (in a non debian format). So far so good. However, I must also read the dependencies (again, of proprietary software within my company) listed in each .deb so as to include them in this "big package". This is the part (reading and extracting the exhaustive list of dependencies) I am not sure how to do and was hoping I can use preexisting debian tools to do this for me. Which one is best or do I have to write this from scratch ? If latter, is there any source I can refer ? I have been at this for a while and need some pointers to preserve my sanity. Thanks for helping !! Daniel Burrows wrote: On Tue, Aug 12, 2008 at 09:05:09PM -0700, NewDeb <[EMAIL PROTECTED]> was heard to say: Well, should'nt 'apt-cache depends' and the control file show the same dependency info ? Here is the inconsistency I am seeing. Please let me know if I am interpreting this wrongly - Control File --- Depends: kaffe (>= 2:1.1.6-3) | java-gcj-compat-dev (>= 1.0.41-1) | java2-runtime, libtomcat5-java (>= 5.0.30-12etch1), adduser (>= 3.34), apache2-utils | apache-utils | apache2-common Suggests: java-virtual-machine, libapache-mod-jk | libapache2-mod-jk, tomcat5-webapps, tomcat5-admin apt-cahce depends tomcat5 (I dont even see "jakarta-commons-launcher" in the above file...) tomcat5-5.5.26-1jpp.2.fc7 ^ You said in your first email on the thread that the control file was for version 5.0.30-12etch1. I imagine that its dependencies are probably different from those of version 5.5.26-1jpp.2.fc7. Daniel -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED] hmm, this is really the problem. For example, Your package tomcat5-5.5.26-1jpp.2.fc7 seems build from fedora core 7, and it depends on jakarta-commons-launcher, this dependence will never satisfied because debian never has this package named jakarta-commons-launcher unless you get this package somewhere and install it. I suggest you install all this debs by dpkg -i *.deb in standard system and then fix the broken dependence by hand. Regards, Ding Honghui -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
How about MySQL in 64bit Debian?
Hi list, How about MySQL server in 64bit Debian? Is it stable enough for product environment? Regards, Ding Honghui -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
gnome-volume-manager, hal and dbus
Hello, I install debian etch, and want my usb disk be mounted automaticly. When I insert my usb disk, the udev find it, and lshal shows it found the device. ps aux shows dbus and gnome-volume-manager are both running. But, the usb device is not automaticly mounted. Any clue to diagnose the problem? I want to know the problem is in dbus or gnome-volume-manager, should dbus-monitor help on that? Regards, Ding Honghui -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
[Resolved] Re: gnome-volume-manager, hal and dbus
hhding.gnu wrote: > Hello, > > I install debian etch, and want my usb disk be mounted automaticly. > > When I insert my usb disk, the udev find it, and lshal shows it found > the device. ps aux shows dbus and gnome-volume-manager are both running. > But, the usb device is not automaticly mounted. > > Any clue to diagnose the problem? > I want to know the problem is in dbus or gnome-volume-manager, should > dbus-monitor help on that? > > Regards, > Ding Honghui > Hello, I find out why it does not works. Because the user start the gnome is not in group plugdev. So dbus-monitor will not get message from dbus. :-) Regards, Ding Honghui -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: device mapper help
William Thompson wrote: > I'm not on the list. Please always keep me in the CC. > > Is there a tool somewhere that will allow me to merge a snapshot into the > parent device? > > For instance, if I have a snapshot of /dev/sdb like this: > sdb-snap1: 0 16777216 snapshot /dev/sdb /dev/sdb-snap1-backing P 8 > sdb-snap2: 0 16777216 snapshot /dev/sdb-snap1 /dev/sdb-snap2-backing P 8 > ... > > Lets say that I want to update sdb with the data in /dev/sdb-snap1-backing. > What would I do to do this? Is there some tool that does this? I > understand that there can be problems with snapshots of snapshots like this, > but that is not the question. > > As far as I know, LVM does not provide this function. ZFS can promote it's snapshoot. Regards, Ding Honghui -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Can expect deal with dialog generated by whiptail?
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 I have some program which show some dialog by whiptail as UI to users, can I use expect to run the program without any interaction? -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.7 (MingW32) iD8DBQFHMrxlJo9Njjkvy34RAo0gAJ9WsaqnJcUk1+NQSn9w5sX7+8D+LgCeJnY6 3TCZEBaHi2lhRI0pHABxqqE= =d/N4 -END PGP SIGNATURE- -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
