Bug 490298

Summary:	When setting max_lv for VG, LV creation can fail and produce invalid metadata
Product:	Red Hat Enterprise Linux 5	Reporter:	Milan Broz <mbroz>
Component:	lvm2	Assignee:	Milan Broz <mbroz>
Status:	CLOSED ERRATA	QA Contact:	Cluster QE <mspqa-list>
Severity:	high	Docs Contact:
Priority:	high
Version:	5.3	CC:	agk, cmarthal, dwysocha, edamato, heinzm, jbrassow, mbroz, prockai, pvrabec
Target Milestone:	rc
Target Release:	---
Hardware:	All
OS:	Linux
Whiteboard:
Fixed In Version:		Doc Type:	Bug Fix
Doc Text:		Story Points:	---
Clone Of:
Clones:	490417 (view as bug list)		Environment:
Last Closed:	2009-09-02 11:56:30 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 Milan Broz 2009-03-14 22:48:07 UTC

Description of problem:

1) Internal VG lv_count is inconsistent is some cases and leads to problems.

# pvcreate /dev/sdc
Physical volume "/dev/sdc" successfully created

# vgcreate -l 4 vg_test /dev/sdc
Volume group "vg_test" successfully created

# lvcreate -L 12M -n lv1 vg_test
Logical volume "lv1" created

# lvcreate -s -L 12M -n lv1s vg_test/lv1
Logical volume "lv1s" created

# lvcreate -L 12M -n lv2 vg_test
Maximum number of logical volumes (4) reached in volume group vg_test
Couldn't read all logical volumes for volume group vg_test.
Maximum number of logical volumes (4) reached in volume group vg_test
Couldn't read all logical volumes for volume group vg_test.
Maximum number of logical volumes (4) reached in volume group vg_test
Couldn't read all logical volumes for volume group vg_test.
Maximum number of logical volumes (4) reached in volume group vg_test
Couldn't read all logical volumes for volume group vg_test.
Volume group for uuid not found: SilYJpDvKDfx2PxlTQ7Ww0bF2Vrte0qRNRtE2LUe9OaHop7J9GPU5CpZ7L3dZ24A
Aborting. Failed to activate new LV to wipe the start of it.
Maximum number of logical volumes (4) reached in volume group vg_test
Couldn't read all logical volumes for volume group vg_test.
Maximum number of logical volumes (4) reached in volume group vg_test
Couldn't read all logical volumes for volume group vg_test.
Maximum number of logical volumes (4) reached in volume group vg_test
Couldn't read all logical volumes for volume group vg_test.
Maximum number of logical volumes (4) reached in volume group vg_test
Couldn't read all logical volumes for volume group vg_test.
Volume group for uuid not found: SilYJpDvKDfx2PxlTQ7Ww0bF2Vrte0qRNRtE2LUe9OaHop7J9GPU5CpZ7L3dZ24A
Unable to deactivate failed new LV. Manual intervention required.

# vgs
Maximum number of logical volumes (4) reached in volume group vg_test
Couldn't read all logical volumes for volume group vg_test.
Maximum number of logical volumes (4) reached in volume group vg_test
Couldn't read all logical volumes for volume group vg_test.
Volume group "vg_test" not found

vg_test is now lost!

2) Moreover, system allows vgcfgrestore with max_lv set to lower value than
actual count of LV. After that, it ignores max_lv completely:

# vgcfgrestore -f vg_test vg_test # with max_lv set to 1
Restored volume group vg_test

# vgs -o +lv_count,max_lv vg_test
VG #PV #LV #SN Attr VSize VFree #LV MaxLV
vg_test 1 2 0 wz--n- 298.09G 298.07G 2 1

# lvcreate -L 12M -n lv2 vg_test
Logical volume "lv2" created

Version-Release number of selected component (if applicable):
lvm2-2.02.40-6.el5.x86_64

Comment 2 Milan Broz 2009-05-19 11:00:55 UTC

The max_lv handling logic is fixed upstream now -> POST

User is not allowed to create new LV if LV count is above limit, but internally tools can temporarily violate this limit (e.g. during mirror conversion).

(Actually max_lv limit make no much sense with lvm2 metadata.)

Comment 3 Milan Broz 2009-05-21 09:22:19 UTC

Fix in version lvm2-2.02.46-1.el5.

Comment 5 Corey Marthaler 2009-07-01 16:48:57 UTC

Fix verified in lvm2-2.02.46-8.el5.

[root@grant-01 ~]# vgcreate -l 4 grant /dev/sd[bc][123]
  Volume group "grant" successfully created
[root@grant-01 ~]# lvcreate -L 12M -n lv1 grant
  Logical volume "lv1" created
[root@grant-01 ~]# lvcreate -s -L 12M -n lv1s grant/lv1
  Logical volume "lv1s" created
[root@grant-01 ~]# lvs -a -o +devices
  LV       VG         Attr   LSize  Origin Snap%  Move Log Copy%  Convert Devices        
  LogVol00 VolGroup00 -wi-ao 64.56G                                       /dev/sda2(0)   
  LogVol01 VolGroup00 -wi-ao  9.81G                                       /dev/sda2(2066)
  lv1      grant      owi-a- 12.00M                                       /dev/sdb1(0)   
  lv1s     grant      swi-a- 12.00M lv1      0.07                         /dev/sdb2(0)   
[root@grant-01 ~]# lvcreate -L 12M -n lv2 grant
  Logical volume "lv2" created
[root@grant-01 ~]# vgs
  VG         #PV #LV #SN Attr   VSize   VFree  
  VolGroup00   1   2   0 wz--n-  74.38G      0 
  grant        6   3   1 wz--n- 544.92G 544.89G
[root@grant-01 ~]# lvcreate -m 1 -n mirror -L 10M grant
  Rounding up size to full physical extent 12.00 MB
  Logical volume "mirror" created
[root@grant-01 ~]# vgs
  VG         #PV #LV #SN Attr   VSize   VFree  
  VolGroup00   1   2   0 wz--n-  74.38G      0 
  grant        6   4   1 wz--n- 544.92G 544.86G
[root@grant-01 ~]# lvcreate -L 12M -n lv2 grant
  Logical volume "lv2" already exists in volume group "grant"
[root@grant-01 ~]# lvcreate -L 12M -n lv3 grant
  Maximum number of logical volumes (4) reached in volume group grant

Comment 7 errata-xmlrpc 2009-09-02 11:56:30 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 therefore 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.

http://rhn.redhat.com/errata/RHBA-2009-1393.html