Hi, Just share my experience with my Terastation Pro II Rackmount with Debian, I got this kind of stuck at booting also and I tried to have serial access to debug. I found that Debian run the disk checks if there's something wrong with the rootfs partition during last boot, I just modify the "pass" parameter in fstab to 0, then the rackmount boot fine again (it took more than 1 hour for 1.5TB HDD checking).
If you setup any pass parameter in fstab, just set it to 0 (the last parameter), then try to boot it again, maybe it will work again. Moreover, I have some spare USB serial cable being made from DKU-5 Nokia one, you really need one? Hope this helps and sorry for my English. Rgds, DH On Thu, Sep 23, 2010 at 4:58 AM, Sylvain L. Sauvage < [email protected]> wrote: > Hi all, > > I got a Qnap TS-210 last week. I installed Debian Sid on it > (with 2x2 TB hdd, RAID-1, ext4) and all was good (rebooted it 3 > or 4 times, copied GBs of data on it, etc.) until Monday evening > when I asked it to reboot and it got stuck (red/green flashing > status light, no disk activity). I tried quite a lot of things > but I don't have a serial cable. > > The short version: > It can boot in recovery mode (TFTP) with the Debian Installer > Testing image. So I can have an ssh console/rescue tool. > It can't boot on any installed Debian (old one, new one, > Testing or Sid, ext3 or ext4 (hmm, I realize I always kept the > RAID-1 though)). > > The longer one: > First, I got the disks out and checked them: they are ok. > (Well, I realized (too long) after I could use the installer as > a console/rescue tool. I would have spent less time moving my > disks around. Well, I do stupid things like that sometimes.) > The logs: nothing, it stops logging in /var/log/dmesg after a > few lines. Other files are empty (or with no more lines than > before the reboot). > I introduced echo lines in some boot scripts (actually, mostly > in /etc/default/*) to easily track where it stopped. I thus > realized that: > 1. It can't write content to files. It can just create them. > (That's weird to me.) > 2. It begins executing rcS.d scripts: it loads > /etc/default/{console-setup, devpts, ifupdown, keyboard, > locale, rcS, tmpfs}. As it doesn't load bootlogd nor > mdadm, I conclude that it stops very early and maybe waits > for some input. > At that stage, I tried reinstalling anew. The installation(s) > went well, but the installed system didn't boot either. (First > with ext4 again, then with only the / partition, in ext3, as I > thought maybe because it can create files but not populate them > then it may be related to the file system. Testing or Sid (but > they have the same kernels so no real difference).) > Then, I restored my not-working-anymore-installation (the disk > part, I did a tarball the first time I got the disks out) and > tried making a TFTP boot image for it with an older linux-image > (2.6.32-5-21 instead of -23). I constructed the initrd from the > 2.6.32-5-23 one, replacing only the modules. It boots as the > others: few seconds disk activity then silence and still > flashing status, no net. > Then, I installed linux-{image,base}-2.6.35-trunk on my > installed system from a chroot from the ssh console of the > installer (I had to force some little things in order to avoid > error or warning messages). Same results. > > What I haven't tried: > Reinstalling the Qnap boot image (not really happy with that). > Not using RAID-1. > > Any idea of anything else I can try? > > By the way, I'm also looking for a TTL-serial converter (with > the good connectors or adapters) in France or Europe (French or > English, and Euros would be better), if you happen to know a > trustworthy (and not too expensive) retailer... > > -- > Sylvain Sauvage > > > -- > To UNSUBSCRIBE, email to [email protected] > with a subject of "unsubscribe". Trouble? Contact > [email protected] > Archive: > http://lists.debian.org/[email protected] > >
