Bug 236075

Summary:	stat failure errors while doing cmirror creation/deletion operations
Product:	[Retired] Red Hat Cluster Suite	Reporter:	Corey Marthaler <cmarthal>
Component:	lvm2-cluster	Assignee:	Milan Broz <mbroz>
Status:	CLOSED CURRENTRELEASE	QA Contact:	Cluster QE <mspqa-list>
Severity:	medium	Docs Contact:
Priority:	medium
Version:	4	CC:	agk, dwysocha, jbrassow, mbroz, prockai, pvrabec
Target Milestone:	---
Target Release:	---
Hardware:	All
OS:	Linux
Whiteboard:
Fixed In Version:		Doc Type:	Bug Fix
Doc Text:		Story Points:	---
Clone Of:		Environment:
Last Closed:	2010-05-17 12:25:52 UTC	Type:	---
Regression:	---	Mount Type:	---
Documentation:	---	CRM:
Verified Versions:		Category:	---
oVirt Team:	---	RHEL 7.3 requirements from Atomic Host:
Cloudforms Team:	---	Target Upstream Version:
Embargoed:

Description Corey Marthaler 2007-04-11 20:10:01 UTC

Description of problem:
In an attempt to verify the fix for bz 213754, I was doing looping create/delete
operations from all 4 nodes in my cluster. This caused some of these operations
to report different stat/lstat errors, even though the operations succeeded.
Also, this may be an lvm2-cluster issue and not necessarily cmirror specific.

Is this happening besause of the deletes on other nodes during the creation process?


[root@link-08 ~]# lvcreate -m 1 -n $(hostname).1 -L500M feist
  /dev/dm-6: stat failed: No such file or directory
  Path /dev/dm-6 no longer valid for device(253,6)
  /dev/mapper/feist-link--02.2_mlog: stat failed: No such file or directory
  Path /dev/mapper/feist-link--02.2_mlog no longer valid for device(253,6)
  Aborting - please provide new pathname for what used to be
/dev/mapper/feist-link--02.2_mlog
  /dev/dm-7: stat failed: No such file or directory
  Path /dev/dm-7 no longer valid for device(253,7)
  /dev/mapper/feist-link--02.2_mimage_0: stat failed: No such file or directory
  Path /dev/mapper/feist-link--02.2_mimage_0 no longer valid for device(253,7)
  Aborting - please provide new pathname for what used to be
/dev/mapper/feist-link--02.2_mimage_0
  /dev/dm-8: stat failed: No such file or directory
  Path /dev/dm-8 no longer valid for device(253,8)
  /dev/mapper/feist-link--02.2_mimage_1: stat failed: No such file or directory
  Path /dev/mapper/feist-link--02.2_mimage_1 no longer valid for device(253,8)
  Aborting - please provide new pathname for what used to be
/dev/mapper/feist-link--02.2_mimage_1
  /dev/dm-9: stat failed: No such file or directory
  Path /dev/dm-9 no longer valid for device(253,9)
  /dev/mapper/feist-link--02.2_mimage_2: stat failed: No such file or directory
  Path /dev/mapper/feist-link--02.2_mimage_2 no longer valid for device(253,9)
  Aborting - please provide new pathname for what used to be
/dev/mapper/feist-link--02.2_mimage_2
  /dev/dm-10: stat failed: No such file or directory
  Path /dev/dm-10 no longer valid for device(253,10)
  /dev/feist/link-02.2: stat failed: No such file or directory
  Path /dev/feist/link-02.2 no longer valid for device(253,10)
  /dev/mapper/feist-link--02.2: stat failed: No such file or directory
  Path /dev/mapper/feist-link--02.2 no longer valid for device(253,10)
  Aborting - please provide new pathname for what used to be
/dev/mapper/feist-link--02.2
  /dev/mapper/feist-link--02.2_mlog: lstat failed: No such file or directory
  /dev/mapper/feist-link--02.2_mimage_2: lstat failed: No such file or directory
  Logical volume "link-08.1" created



Version-Release number of selected component (if applicable):
[root@link-08 ~]# rpm -q lvm2
lvm2-2.02.21-3.el4
[root@link-08 ~]# rpm -q lvm2-cluster
lvm2-cluster-2.02.21-3.el4
[root@link-08 ~]# rpm -q cmirror-kernel
cmirror-kernel-2.6.9-30.0

Comment 1 Jonathan Earl Brassow 2007-04-19 14:49:27 UTC

could we get a -vvvv output for this?

Comment 3 Jonathan Earl Brassow 2008-03-26 18:25:17 UTC

I see this often as well.  It appears that device-mapper devices are being put
in the LVM cache - then LVM commands complain when it doesn't find them (because
another machine removed them while the lvm command was in progress?).

Comment 6 Milan Broz 2010-05-17 12:25:52 UTC

I think there were so many changes to this code, some of them surely fixes at least part of this... 

If you still see this, please reopen, thanks.