Bug 738174

Summary: ms ext4: ext3->ext4 migration is broken
Product: Red Hat Enterprise Linux 6 Reporter: Monakhov Dmitriy <dmonakhov>
Component: kernelAssignee: fs-maint
Status: CLOSED WONTFIX QA Contact: Filesystem QE <fs-qe>
Severity: high Docs Contact:
Priority: unspecified    
Version: 6.3CC: eguan, esandeen, rwheeler
Target Milestone: rc   
Target Release: ---   
Hardware: Unspecified   
OS: Linux   
Whiteboard:
Fixed In Version: Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2013-03-24 20:00:47 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:
Attachments:
Description Flags
testcase for the bug none

Description Monakhov Dmitriy 2011-09-14 09:04:41 UTC 
Created attachment 523092 [details]
testcase for the bug

This is mainstream bug.
it happens during migration inode from ind-layout to extent one
chattr +e ./a_file and result in massive data corruption.
Patch already available: http://patchwork.ozlabs.org/patch/93790/
and it is applicable to 2.6.32.131 as is.
Test case attached.
Comment 2 RHEL Program Management 2011-10-07 15:48:10 UTC 
Since RHEL 6.2 External Beta has begun, and this bug remains
unresolved, it has been rejected as it is not proposed as
exception or blocker.

Red Hat invites you to ask your support representative to
propose this request, if appropriate and relevant, in the
next release of Red Hat Enterprise Linux.
Comment 3 Eric Sandeen 2012-10-05 14:23:50 UTC 
Just a note - we do not support ext3->ext4 migration in RHEL (as is probably obvious by the fact that this stuff is so broken).

Migration paths are rarely run or tested, and never run or tested by our internal QE.  Migration gets you half-an-ext4, without all of the improved metadata layout you gain at mkfs time.  Our only supported usecase for ext4 is a fresh mkfs.

Since this is simple & obvious, we could consider fixing it to avoid corruption for people who try it anyway, but I don't want to leave the impression that this is a supported/maintained use case.
Comment 4 Eric Sandeen 2013-03-24 20:00:47 UTC 
Since migration isn't supported in RHEL, I'm just going to WONTFIX this one.

Thanks for the report,
-Eric