[Ocfs2-users] very slow writes

[Jay V]

Ever since the file system reached 82% full, the file system has gone to a
crawl and stalls of only writing up to 10 mb/s, even going as slow as a few
kbytes/sec. The file system may be fragmented. Is there a defrag tool?

I have tried many options, but nothing seems to work. "data=writeback" does
seem to work. We have checked the logs, debugs, traces trying to figure out
the probelm, but with no success.

Features:
Compat: backup-super strict-journal-super
Incompat: sparse inline-data
RO Compat: unwritten
debugfs.ocfs2 -R "stats" /dev/drbd2
        Revision: 0.90
        Mount Count: 0   Max Mount Count: 20
        State: 0   Errors: 0
        Check Interval: 0   Last Check: Sat Mar 17 05:51:42 2012
        Creator OS: 0
        Feature Compat: 3 backup-super strict-journal-super
        Feature Incompat: 80 sparse inline-data
        Tunefs Incomplete: 0
        Feature RO compat: 1 unwritten
        Root Blknum: 5   System Dir Blknum: 6
        First Cluster Group Blknum: 3
        Block Size Bits: 12   Cluster Size Bits: 12
        Max Node Slots: 2
        Extended Attributes Inline Size: 0
        Label: imepool2
        UUID: 5AAD4275BEBA4F96B8FEE7274FC4D043
        Hash: 0 (0x0)
        DX Seeds: 0 0 0 (0x00000000 0x00000000 0x00000000)
        Cluster stack: classic o2cb
        Cluster flags: 0
        Inode: 2   Mode: 00   Generation: 2084865048 (0x7c448418)
        FS Generation: 2084865048 (0x7c448418)
        CRC32: 00000000   ECC: 0000
        Type: Unknown   Attr: 0x0   Flags: Valid System Superblock
        Dynamic Features: (0x0)
        User: 0 (root)   Group: 0 (root)   Size: 0
        Links: 0   Clusters: 3496426618
        ctime: 0x4aa1e6e7 0x0 -- Fri Sep  4 21:19:51.0 2009
        atime: 0x0 0x0 -- Wed Dec 31 16:00:00.0 1969
        mtime: 0x4aa1e6e7 0x0 -- Fri Sep  4 21:19:51.0 2009
        dtime: 0x0 -- Wed Dec 31 16:00:00 1969
        Refcount Block: 0
        Last Extblk: 0   Orphan Slot: 0
        Sub Alloc Slot: Global   Sub Alloc Bit: 65535

Version OCFS2 1.8.0:
filename:
/lib/modules/2.6.39-100.5.1.el6uek.x86_64/kernel/fs/ocfs2/ocfs2.ko
license:        GPL
author:         Oracle
version:        1.8.0
description:    OCFS2 1.8.0
srcversion:     A427F5E2A962D0808C8478F
depends:        jbd2,ocfs2_stackglue,ocfs2_nodemanager
vermagic:       2.6.39-100.5.1.el6uek.x86_64 SMP mod_unload modversions
mounted options: rw,_netdev,noatime,nodiratime,heartbeat=local

When deleting files it seems to help temporarily. I like to be able to go
over 82% without this degradation.

I want to minimize downtime as this filesystem is 15TB. Is there a feature
I can enable such as "discontig-bg" that will help??

Thanks so much for your help!
-Jay
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://oss.oracle.com/pipermail/ocfs2-users/attachments/20120510/091763f4/attachment.html