Date: Sat, 27 Mar 2004 20:30:46 -0800 (PST) From: Drew Tomlinson <nojunk4me2004@yahoo.com> To: Greg 'groggy' Lehey <grog@FreeBSD.org> Cc: freebsd-questions@freebsd.org Subject: Re: Vinum Subdisk Crashed - How To Repair? Message-ID: <20040328043046.46103.qmail@web41712.mail.yahoo.com> In-Reply-To: <20040327063358.GG52771@wantadilla.lemis.com>
next in thread | previous in thread | raw e-mail | index | archive | help
--- Greg 'groggy' Lehey <grog@FreeBSD.org> wrote: > On Friday, 26 March 2004 at 21:53:44 -0800, Drew > Tomlinson wrote: > > Please excuse whatever format in which this email > > arrives. My system is unusable so I am posting > > from Yahoo!. Thank you very much for your reply. I’ve tried to clean this up so it remains readable when it reaches you. > Heh. I suppose that's a good enough reason. > > I have a 4.9-RELEASE-p4 system. I have not made > > any changes to any source code. I have 3 vinum > > volumes configured. While attempting to diagnose > > problems with one of the volumes that uses a > > firewire drive, my system crashed with a trap > > 12 error. I have my /usr configured as a > > striped vinum volume with one plex and two > > subdisks. Upon reboot, the usr volume has one > > subdisk in a > > crashed state. Thus I can't mount /usr and > > can't get my system up and running. > > > > I *REALLY* don't want to lose my data on /usr > > as I don't have a backup. It was my backup > > volume problems I was looking at when my > > trouble started. I have looked at > > http://www.vinumvm.org/vinum/how-to-debug.html. I > > have a copy of my config files and the page > > says one can use those to recreate his > > config without losing data as it does not write to > > the subdisks. Is this what I should do? > > Maybe. > > > Should I do something else? > > Maybe. > > > Please cc me on any replies as I am not able to > > receive them via my subscription at the moment. > > > I can not provide the extract of > > /var/log/vinum_history or /var/log/messages as > > I symlinked /var to /usr/var. > > > > vinum -> list > > 4 drives: > > D ftp1 State: up Device /dev/ad0s1h Avail: 76318/76319 MB (100%) > > D ftp2 State: up Device /dev/ad1s1h Avail: 0/76319 MB (0%) > > D disk1 State: up Device /dev/da0s1h Avail: 84701/8383 MB (1010%) > > This should go away if you stop and restart Vinum. > > D disk2 State: up Device /dev/da1s1h Avail: 0/8383 MB (0%) > > D *invalid* State: referenced Device Avail: 0/0 MB > > This one looks dangerous. This is what I was trying to diagnose when the system crashed and I lost /usr. The device is /dev/da2s1h and is an IDE drive in a firewire enclosure. It appears that upon reboot, the drive is not recognized by the system as it is not in the dmesg output. Once the system has completed its reboot, a simple unplug/plug of the firewire cable resets the firewire bus and then the drive is found until next reboot. There is probably a way to do this via the command line but I haven't found it yet. > > 3 volumes: > > V usr State: up Plexes: 1 Size: 16 GB > > V ftp State: up Plexes: 1 Size: 149 GB > > V backup State: down Plexes: 1 Size: 74 GB > > > > 3 plexes: > > P usr.p0 S State: corrupt Subdisks: 2 Size: 16 GB > > P ftp.p0 C State: up Subdisks: 2 Size: 149 GB > > P backup.p0 C State: faulty Subdisks: 1 Size: 74 GB > > > > 5 subdisks: > > S usr.p0.s0 State: up PO: 0 B Size: 8383 MB > > S usr.p0.s1 State: crashed PO: 256 kB Size: 8383 MB > > S ftp.p0.s0 State: up PO: 0 B Size: 74 GB > > S backup.p0.s0 State: crashed PO: 0 B Size: 74 GB > > S ftp.p0.s1 State: up PO: 74 GB Size: 74 GB > > It looks to me as if you have lost a drive. > Hopefully this is simple, > but the drive on which usr.p0.s1 and backup.p0.s0 > are located is presumably the > one which is referenced without a name. backup.p0.s0 is on the unreferenced drive. usr.p0.s0 is on /dev/da1s1h as defined in the 'disk2' drive. Because I am still in the testing and setup phase with Bacula, I can lose everything related to my backup volume. OTOH, /usr is a different story. :) > Before going too much further, > > 1. Stop and restart Vinum and see if the > availability statistics > return to "normal". If not, reply with the > values, even if > they're the same as above. I tried this but the problem remains. I had to use 'setdaemon 1' to get 'stop' to have any effect. Here's the output: vinum -> stop *** Warning: configuration updates are disabled. *** vinum -> setdaemon 1 vinum -> stop vinum: stopping vinum: unloaded vinum unloaded # vinum vinum: loaded vinum -> l 0 drives: 0 volumes: 0 plexes: 0 subdisks: vinum -> start vinum: reading configuration from /dev/da1s1h vinum: usr.p0.s1 is crashed vinum: usr.p0 is corrupt vinum: No space for on disk1 Disabling configuration updates vinum: backup.p0.s0 is crashed vinum: backup.p0 is faulty vinum: backup is down vinum: updating configuration from /dev/ad1s1h vinum: updating configuration from /dev/da0s1h vinum: updating configuration from /dev/ad0s1h Correcting length of ftp.p0: was 468903669, is 312602446 *** Warning: configuration updates are disabled. *** Warning: defective objects D *invalid* State: referenced Device Avail: 0/0 MB V backup State: down Plexes: 1 Size: 74 GB P usr.p0 S State: corrupt Subdisks: 2 Size: 16 GB P backup.p0 C State: faulty Subdisks: 1 Size: 74 GB S usr.p0.s1 State: crashed PO: 256 kB Size: 8383 MB S backup.p0.s0 State: crashed PO: 0 B Size: 74 GB *** Warning: configuration updates are disabled. *** vinum -> list 4 drives: D ftp1 State: up Device /dev/ad0s1h Avail: 76318/76319 MB (100%) D ftp2 State: up Device /dev/ad1s1h Avail: 0/76319 MB (0%) D disk1 State: up Device /dev/da0s1h Avail: 84701/8383 MB (1010%) D disk2 State: up Device /dev/da1s1h Avail: 0/8383 MB (0%) D *invalid* State: referenced Device Avail: 0/0 MB 3 volumes: V usr State: up Plexes: 1 Size: 16 GB V ftp State: up Plexes: 1 Size: 149 GB V backup State: down Plexes: 1 Size: 74 GB 3 plexes: P usr.p0 S State: corrupt Subdisks: 2 Size: 16 GB P ftp.p0 C State: up Subdisks: 2 Size: 149 GB P backup.p0 C State: faulty Subdisks: 1 Size: 74 GB 5 subdisks: S usr.p0.s0 State: up PO: 0 B Size: 8383 MB S usr.p0.s1 State: crashed PO: 256 kB Size: 8383 MB S ftp.p0.s0 State: up PO: 0 B Size: 74 GB S backup.p0.s0 State: crashed PO: 0 B Size: 74 GB S ftp.p0.s1 State: up PO: 74 GB Size: 74 GB > 2. Where are usr.p0.s1 and backup.p0.s0 located? use.p0.s1 is on /dev/da1s1h and backup.p0.s0 is on /dev/da2s1h. > If they're on one > of the drives in the list, you could try some > selective deletion. > If this is the case, I'll give you more details. > > It's probably worth posting the configuration file. Because I built my vinum volumes at separate times and since the configuration is cumulative, I have 3 files, one for each volume. Here they are: --- usr volume # cat vinum_usr.conf drive disk1 device /dev/da0h drive disk2 device /dev/da1h volume usr plex org striped 256k sd length 0 drive disk1 sd length 0 drive disk2 --- ftp volume # cat vinum_ftp.conf drive ftp1 device /dev/ad0h drive ftp2 device /dev/ad1h volume ftp plex org concat sd length 0 drive ftp1 sd length 0 drive ftp2 --- backup volume # cat vinum_backup.conf drive backup device /dev/da2h volume backup plex org concat sd length 0 drive backup Thank you very much for your help and willingness to work through the poor formatting of my emails. Drew __________________________________ Do you Yahoo!? Yahoo! Finance Tax Center - File online. File on time. http://taxes.yahoo.com/filing.html
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20040328043046.46103.qmail>