[Bug 105936] Re: snapshot creation failure race in use: not deactivating

2009-11-09 Thread Scott James Remnant
** Changed in: lvm2 (Ubuntu) Assignee: Scott James Remnant (scott) = (unassigned) -- snapshot creation failure race in use: not deactivating https://bugs.launchpad.net/bugs/105936 You received this bug notification because you are a member of Ubuntu Bugs, which is subscribed to Ubuntu. --

[Bug 105936] Re: snapshot creation failure race in use: not deactivating

2008-09-23 Thread Scott James Remnant
The fix is likely to be invasive, so not eligible for packporting - especially past an LTS release ** Changed in: lvm2 (Ubuntu Gutsy) Status: Confirmed = Won't Fix -- snapshot creation failure race in use: not deactivating https://bugs.launchpad.net/bugs/105936 You received this bug

[Bug 105936] Re: snapshot creation failure race in use: not deactivating

2007-09-05 Thread Alex Muntada
FWIW, I don't know if this bug isn't reproducible in Gutsy (I won't upgrade my servers until Gutsy is released) but it definitely is reproducible in Feisty with Kees' latest test script: # bash snapshot-collision.sh xen 3 5 512M 256M Dropping disk caches ... Creating test LVs ... Logical volume

[Bug 105936] Re: snapshot creation failure race in use: not deactivating

2007-09-05 Thread Kees Cook
I have not been able to reproduce it on gutsy. As Scott says, it is technically still possible, but it is much much harder to run into on accident now. Feisty's situation is not easily fixed. -- snapshot creation failure race in use: not deactivating https://bugs.launchpad.net/bugs/105936 You

[Bug 105936] Re: snapshot creation failure race in use: not deactivating

2007-09-04 Thread Scott James Remnant
Removing milestone; this is merely a theoretical bug and should not be demonstrable in the field. The only way to replicate it is by gdb'ing the kernel (through vmware) and mucking around -- we keep it open because there is a race here, and that should be fixed ** Changed in: lvm2 (Ubuntu

[Bug 105936] Re: snapshot creation failure race in use: not deactivating

2007-08-31 Thread Henrik Nilsen Omma
Moving milestone. ** Changed in: lvm2 (Ubuntu Gutsy) Target: tribe-5 = tribe-6 -- snapshot creation failure race in use: not deactivating https://bugs.launchpad.net/bugs/105936 You received this bug notification because you are a member of Ubuntu Bugs, which is the bug contact for

[Bug 105936] Re: snapshot creation failure race in use: not deactivating

2007-08-27 Thread Joe Kislo
So I've been struggling with LVM snapshot issues for atleast a year or so (in some form or another) I'm not sure if the kernel engineers working on this issue have nailed down the precise cause, but I have made some observations over the past week that may help. 1) The issue is has a definite

[Bug 105936] Re: snapshot creation failure race in use: not deactivating

2007-08-08 Thread Martin Pitt
** Changed in: lvm2 (Ubuntu Gutsy) Target: tribe-4 = tribe-5 -- snapshot creation failure race in use: not deactivating https://bugs.launchpad.net/bugs/105936 You received this bug notification because you are a member of Ubuntu Bugs, which is the bug contact for Ubuntu. -- ubuntu-bugs

[Bug 105936] Re: snapshot creation failure race in use: not deactivating

2007-07-31 Thread Darik Horn
I've got a computer that always expresses this bug, perhaps because the filesystem is XFS. Let me know if I can help test a fix. -- snapshot creation failure race in use: not deactivating https://bugs.launchpad.net/bugs/105936 You received this bug notification because you are a member of

[Bug 105936] Re: snapshot creation failure race in use: not deactivating

2007-07-31 Thread Darik Horn
My comment about XFS was mistaken. The glitch can be reliably reproduced only because I am trying to create very large snapshots. I'm getting the same race condition with small snapshots. -- snapshot creation failure race in use: not deactivating https://bugs.launchpad.net/bugs/105936 You

[Bug 105936] Re: snapshot creation failure race in use: not deactivating

2007-07-17 Thread Martin Pitt
Does not look like it is getting fixed in tijme for tribe-3, moving. ** Changed in: lvm2 (Ubuntu Gutsy) Target: tribe-3 = tribe-4 -- snapshot creation failure race in use: not deactivating https://bugs.launchpad.net/bugs/105936 You received this bug notification because you are a member

[Bug 105936] Re: snapshot creation failure race in use: not deactivating

2007-06-22 Thread Martin Pitt
moving to tribe-3 as agreed on yesterday's meeting ** Changed in: lvm2 (Ubuntu Gutsy) Target: tribe-2 = tribe-3 -- snapshot creation failure race in use: not deactivating https://bugs.launchpad.net/bugs/105936 You received this bug notification because you are a member of Ubuntu Bugs,

[Bug 105936] Re: snapshot creation failure race in use: not deactivating

2007-06-08 Thread Kees Cook
Updated test harness to use a more reliable cache-clearing mechanism (thanks shawarma!) ** Attachment added: new test script http://launchpadlibrarian.net/8028958/snapshot-collision.sh ** Attachment removed: test harness for LVM snapshots

[Bug 105936] Re: snapshot creation failure race in use: not deactivating

2007-06-08 Thread Kees Cook
Since the udevmapper changes from Feisty were reverted, I have not seen this bug any more. Does anyone else still see it? ** Changed in: lvm2 (Ubuntu Gutsy) Status: Confirmed = Needs Info -- snapshot creation failure race in use: not deactivating https://bugs.launchpad.net/bugs/105936

[Bug 105936] Re: snapshot creation failure race in use: not deactivating

2007-06-08 Thread Kees Cook
** Changed in: lvm2 (Ubuntu Gutsy) Assignee: (unassigned) = Scott James Remnant Status: Needs Info = Confirmed -- snapshot creation failure race in use: not deactivating https://bugs.launchpad.net/bugs/105936 You received this bug notification because you are a member of Ubuntu Bugs,

[Bug 105936] Re: snapshot creation failure race in use: not deactivating

2007-06-08 Thread Craig Box
Is this fix going to be backportable to Feisty? -- snapshot creation failure race in use: not deactivating https://bugs.launchpad.net/bugs/105936 You received this bug notification because you are a member of Ubuntu Bugs, which is the bug contact for Ubuntu. -- ubuntu-bugs mailing list

[Bug 105936] Re: snapshot creation failure race in use: not deactivating

2007-06-07 Thread Joe Kislo
I can confirm this happens on feisty AMD64. Linux db-1-snapshot 2.6.20-16-server #2 SMP Wed May 23 00:36:09 UTC 2007 x86_64 GNU/Linux db-1-snapshot - error creating snapshot lvcreate:Setting chunksize to 16 sectors. Finding volume group vg_san_mysql Archiving volume group

[Bug 105936] Re: snapshot creation failure race in use: not deactivating

2007-06-07 Thread Alex Muntada
@Joe Kislo: lvcreate reports an error in $? when snapshot fails, but just retrying won't work since you'll need to remove the bad LV first because it is actually created. However, you can check it out in your scripts by looking at lvdisplay flags for that LV (e.g. see the attachment included).

[Bug 105936] Re: snapshot creation failure race in use: not deactivating

2007-06-05 Thread Scott James Remnant
** Changed in: lvm2 (Ubuntu Gutsy) Target: None = tribe-2 -- snapshot creation failure race in use: not deactivating https://bugs.launchpad.net/bugs/105936 You received this bug notification because you are a member of Ubuntu Bugs, which is the bug contact for Ubuntu. -- ubuntu-bugs

[Bug 105936] Re: snapshot creation failure race in use: not deactivating

2007-05-25 Thread Alex Muntada
Usually after performing a lvremove and retrying lvcreate again works. BTW, Kees, were you using a xen kernel too? -- snapshot creation failure race in use: not deactivating https://bugs.launchpad.net/bugs/105936 You received this bug notification because you are a member of Ubuntu Bugs, which

[Bug 105936] Re: snapshot creation failure race in use: not deactivating

2007-05-25 Thread Alex Muntada
I'm confirming this issue in Feisty's xen kernel 2.6.19-4-server: # ./snapshot-collision.sh xen 4 1 4G 512M Attempting to empty disk caches ... 1024+0 records in 1024+0 records out 1073741824 bytes (1.1 GB) copied, 6.87661 seconds, 156 MB/s Creating test LVs ... Logical volume poc1 created

[Bug 105936] Re: snapshot creation failure race in use: not deactivating

2007-05-25 Thread Kees Cook
Yeah, there is a race condition with the udev helper tools and the lvm tools while the support devices are being created/removed. I was using a stock Feisty kernel, non-Xen. (And this is still present in Gutsy too). ** Changed in: lvm2 (Ubuntu) Importance: Undecided = High -- snapshot

[Bug 105936] Re: snapshot creation failure race in use: not deactivating

2007-05-03 Thread Craig Box
Same problem is reoccuring - when lvcreate has run, the volume is created, but not marked as a snapshot volume. LV VG Attr LSizeOrigin Snap% Move Log Copy% 2007.04.30-19.00.00 mlc2 swi-ao3.91G storage 10.32 2007.05.01-00.00.00 mlc2 -wi-a-3.42G

[Bug 105936] Re: snapshot creation failure race in use: not deactivating

2007-05-03 Thread Craig Box
Debug output: [EMAIL PROTECTED]:/etc/udev# lvcreate -L500M -s -n test /dev/mlc2/storage -v Setting chunksize to 16 sectors. Finding volume group mlc2 Archiving volume group mlc2 metadata (seqno 261). Creating logical volume test Creating volume group backup /etc/lvm/backup/mlc2

[Bug 105936] Re: snapshot creation failure race in use: not deactivating

2007-04-27 Thread Craig Box
I had a couple of snaphots on a machine, which I rebooted, and I had to remove before I could mount the underlying LV. Now, I can't create new ones, with the same error as in this bug. -- snapshot creation failure race in use: not deactivating https://bugs.launchpad.net/bugs/105936 You received

[Bug 105936] Re: snapshot creation failure race in use: not deactivating

2007-04-27 Thread Craig Box
Although after a few minutes wait, now I can! (I am very sorry that my bug reports for LVM are so irrational. :-) -- snapshot creation failure race in use: not deactivating https://bugs.launchpad.net/bugs/105936 You received this bug notification because you are a member of Ubuntu Bugs, which

[Bug 105936] Re: snapshot creation failure race in use: not deactivating

2007-04-12 Thread Kees Cook
Here is an example of the attached test script hitting the bug: # ./snapshot-collision.sh systemvg 4 1 4G 2G Attempting to empty disk caches ... 1024+0 records in 1024+0 records out 1073741824 bytes (1.1 GB) copied, 8.92084 seconds, 120 MB/s Creating test LVs ... Logical volume poc1 created