[Cluster-devel] [GFS2 PATCH] data corruption fix
Wendy Cheng
wcheng at redhat.com
Tue Sep 18 13:19:13 UTC 2007
* GFS2 has been using i_cache array to store its indirect meta blocks.
Its flush routine doesn't correctly clean up all the entries. The
problem would show while multiple nodes do simultaneous writes to the
same file. Upon glock exclusive lock transfer, if the file is a sparse
file with large file size where the indirect meta blocks span multiple
array entries with "zero" entries in between. The flush routine
prematurely stops the flushing that leaves old (stale) entries around.
This leads to several nasty issues, including data corruption.
* Fix gfs2_get_block_noalloc checking to correctly return EIO upon
unmapped buffer.
* Simplify gfs2 code path by disabling gfs2_writepages. The
implementation currently generates unmapped dirty buffers with mmap
calls that do not work well with rest of the gfs2 journaling code. Defer
the writepages implementation to later time frame. All writepage
function now goes thru gfs2_writepage call with this change.
-- Wendy
-------------- next part --------------
A non-text attachment was scrubbed...
Name: gfs2_meta_fix.patch
Type: text/x-patch
Size: 1821 bytes
Desc: not available
URL: <http://listman.redhat.com/archives/cluster-devel/attachments/20070918/30c59f88/attachment.bin>
More information about the Cluster-devel
mailing list