Bug 244202 - Boot fails if any LVM snapshots are full
Summary: Boot fails if any LVM snapshots are full
Alias: None
Product: Red Hat Enterprise Linux 5
Classification: Red Hat
Component: mkinitrd
Version: 5.0
Hardware: All
OS: Linux
Target Milestone: ---
: ---
Assignee: Peter Jones
QA Contact: James Laska
Depends On:
TreeView+ depends on / blocked
Reported: 2007-06-14 14:26 UTC by Stephen Tweedie
Modified: 2013-09-02 06:20 UTC (History)
2 users (show)

Clone Of:
Last Closed: 2008-05-21 15:26:19 UTC

Attachments (Terms of Use)
this is a fix. (416 bytes, patch)
2007-06-14 15:09 UTC, Peter Jones
no flags Details | Diff

External Trackers
Tracker ID Priority Status Summary Last Updated
Red Hat Product Errata RHBA-2008:0437 normal SHIPPED_LIVE mkinitrd bug fix and enhancement update 2008-05-20 13:11:44 UTC

Description Stephen Tweedie 2007-06-14 14:26:21 UTC
Description of problem:
When booting off LVM, the initrd tries to activate all LVs.  This fails if there
are any full snapshots, and initrd fails to recover from the error, leading to a
failure to boot.

Version-Release number of selected component (if applicable):

How reproducible:
100%, I believe.

Steps to Reproduce:
1. Create a small snapshot of the root filesystem ("lvcreate -s ...")
2. Write enough data to the root filesystem to overflow the snapshot ("lvs" will
 show it at 100% full)
3. Reboot
Actual results:
Boot fails with a nash SEGV

Expected results:
Should boot!

Comment 1 Stephen Tweedie 2007-06-14 14:27:04 UTC
Boot log from the boot failure:

device-mapper: ioctl: 4.11.0-ioctl (2006-09-14) initialised: dm-devel@redhat.com
  Reading all physical volumes.  This may take a while...
  Found volume group "spectre" using metadata type lvm2
device-mapper: snapshots: snapshot is marked invalid
device-mapper: table: 253:3: snapshot: Failed to read snapshot metadata
device-mapper: ioctl: error adding target to table
  device-mapper: reload ioctl failed: Invalid argument
  10 logical volume(s) in volume group "spectre" now active
nash received SIGSEGV!  Backtrace:
Kernel panic - not syncing: Attempted to kill init!

Comment 2 Stephen Tweedie 2007-06-14 14:28:23 UTC
Fortunately, a readonly rescue boot off CD _does_ boot successfully, and the
situation can then be recovered by removing the full snapshot volume.

Comment 3 Peter Jones 2007-06-14 15:09:38 UTC
Created attachment 157009 [details]
this is a fix.

Fix is attached.  Proposing for 5.2 .

Comment 4 Milan Broz 2007-06-14 16:55:21 UTC
Related to bug 244215 - activation of snapshot marked invalid.

Comment 5 RHEL Product and Program Management 2007-10-16 03:57:39 UTC
This request was evaluated by Red Hat Product Management for inclusion in a Red
Hat Enterprise Linux maintenance release.  Product Management has requested
further review of this request by Red Hat Engineering, for potential
inclusion in a Red Hat Enterprise Linux Update release for currently deployed
products.  This request is not yet committed for inclusion in an Update

Comment 7 Peter Jones 2008-01-31 23:11:05 UTC
Fixed in mkinitrd- .

Comment 9 James Laska 2008-04-11 16:34:22 UTC
Unable to reproduce on 5.1, however 5.0 appears to consistently trigger the failure.

With the help of Stephen, created a 5.0 and 5.2 (snap#5 20080409) install with a
full lvm snapshot, as seen below:

  LV    VG         Attr   LSize  Origin Snap%  Move Log Copy% 
  root  VolGroup00 owi-ao  3.91G                              
  snap1 VolGroup00 Swi-I- 32.00M root   100.00                
  swap  VolGroup00 -wi-ao  1.94G 

 * The i386+x86_64+ppc+s390x 5.0 systems nash segfault on reboot as noted in
 * The i386+x86_64+ppc+s390x 5.2 systems do not segfault upon reboot.

Moving to VERIFIED

Comment 11 errata-xmlrpc 2008-05-21 15:26:19 UTC
An advisory has been issued which should help the problem
described in this bug report. This report is therefore being
closed with a resolution of ERRATA. For more information
on the solution and/or where to find the updated files,
please follow the link below. You may reopen this bug report
if the solution does not work for you.


Note You need to log in before you can comment on or make changes to this bug.