Ext3 fsck performance

[Broekman, Maarten]

I'm having an issue on a couple of my RHEL3 servers that have some
filesystems on them that are pretty large.  These are all ext3
filesystems on SAN storage.  Unfortunately, when we need to fsck these
filesystems, they can take anywhere from 3 to 12 hours each.  We have
similar sized journaled filesystems on AIX and Solaris systems that
don't have these fsck times.  Since ext3 is also journaled, we would
expect the fsck times to be similar.

I was looking at the various journaling modes, but I'm not sure if that
would help the fsck performance necessarily.  In particular, I was
thinking of journaling both the data and the metadata.

Does anyone have any suggestions for performance tuning, mount options,
etc that would improve fsck performance?

Problem filesystems:
/dev/app/appdata0     817G  689G  120G  86% /appdata0
/dev/app/appdata1     635G  571G   58G  91% /appdata1
/dev/app/appdata2     552G  387G  160G  71% /appdata2
/dev/app/appdata3     553G  404G  143G  74% /appdata3

Thanks
Maarten