Bug 455670

Summary:	Mirror to mirror upconvert is broken (especially wrt cluster mirrors)
Product:	Red Hat Enterprise Linux 5	Reporter:	Jonathan Earl Brassow <jbrassow>
Component:	lvm2	Assignee:	Jonathan Earl Brassow <jbrassow>
Status:	CLOSED ERRATA	QA Contact:	Cluster QE <mspqa-list>
Severity:	high	Docs Contact:
Priority:	high
Version:	5.2	CC:	agk, bstevens, cmarthal, dwysocha, edamato, heinzm, jbrassow, lhh, mbroz, prockai
Target Milestone:	rc
Target Release:	---
Hardware:	All
OS:	Linux
Whiteboard:
Fixed In Version:		Doc Type:	Bug Fix
Doc Text:		Story Points:	---
Clone Of:		Environment:
Last Closed:	2009-01-20 21:46:03 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 Jonathan Earl Brassow 2008-07-16 21:43:39 UTC

When going from 2-way mirrors to 3-way mirrors, the 3-way mirror is activated
(resumed in device-mapper) before the 2-way mirror is deactivated
(suspended/removed in device-mapper).

I assume that the way this is happening is that when a new mirror gets stacked
on the first to form the 3-way, there is a point in time when the old 2-way gets
activated twice.

Comment 2 Jonathan Earl Brassow 2008-07-16 22:06:03 UTC

... and patches existed in February, but for some reason, didn't make it into
CVS or rhel...

https://www.redhat.com/archives/lvm-devel/2008-February/msg00007.html

Comment 3 Alasdair Kergon 2008-07-17 01:14:05 UTC

There was discussion at the time concluding that an in-kernel enhancement with
interface extension was the correct long-term solution.

Those patches were not acceptable as presented because they just replaced one
problem with a different problem without justification.

We can only do that if a judgement can be made that, out in the real world,
having the new problem is somehow "less bad" than having the existing problem. 
No such case was made, so I retained the status quo.

Comment 4 Corey Marthaler 2008-09-22 20:46:24 UTC

Upconvert is still broken with the latest code.

2.6.18-116.el5

lvm2-2.02.40-2.el5    BUILT: Fri Sep 19 09:46:26 CDT 2008
lvm2-cluster-2.02.40-2.el5    BUILT: Fri Sep 19 09:49:59 CDT 2008
device-mapper-1.02.28-2.el5    BUILT: Fri Sep 19 02:50:32 CDT 2008
cmirror-1.1.25-1.el5    BUILT: Fri Sep 19 16:27:46 CDT 2008
kmod-cmirror-0.1.17-1.el5    BUILT: Fri Sep 19 16:27:33 CDT 2008

[root@taft-03 ~]# lvconvert -m 2 taft/mirror
[hang]

Sep 22 15:40:22 taft-03 kernel: device-mapper: dm-log-clustered: [rFuXj1gF] Request timed g
device-mapper: dm-log-clustered: [rFuXj1gF] Request timed out: [DM_CLOG_POSTSUSPEND/3491] g
Sep 22 15:40:37 taft-03 kernel: device-mapper: dm-log-clustered: [rFuXj1gF] Request timed g
device-mapper: dm-log-clustered: [rFuXj1gF] Request timed out: [DM_CLOG_POSTSUSPEND/3492] g

Comment 6 Kiersten (Kerri) Anderson 2008-10-20 14:17:26 UTC

Turns out this is related to bug 460156 which is a blocker for 5.3.  Changing flags to 5.3 request and blocker status.

Comment 8 Milan Broz 2008-10-24 12:32:17 UTC

Patch in
lvm2-2.02.40-6.el5.src.rpm
lvm2-cluster-2.02.40-6.el5.src.rpm

Comment 11 errata-xmlrpc 2009-01-20 21:46:03 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-2009-0179.html