Bug 765509 (GLUSTER-3777) - Iozone process hung at 128MB
Summary: Iozone process hung at 128MB
Keywords:
Status: CLOSED WORKSFORME
Alias: GLUSTER-3777
Product: GlusterFS
Classification: Community
Component: unclassified
Version: 3.2.4
Hardware: x86_64
OS: Linux
medium
medium
Target Milestone: ---
Assignee: shishir gowda
QA Contact:
URL:
Whiteboard:
Depends On:
Blocks:
TreeView+ depends on / blocked
 
Reported: 2011-11-03 05:36 UTC by Anush Shetty
Modified: 2013-12-09 01:27 UTC (History)
4 users (show)

Fixed In Version:
Doc Type: Bug Fix
Doc Text:
Clone Of:
Environment:
Last Closed:
Regression: ---
Mount Type: ---
Documentation: ---
CRM:
Verified Versions:

Attachments (Terms of Use)
Add an attachment (proposed patch, testcase, etc.)

Description Anush Shetty 2011-11-03 05:36:44 UTC 
Iozone hangs consistently on RHEL6.1 VM over distributed-replicate setup. This is the dmesg trace. 

Call Trace:
 [<ffffffff8110d3d0>] ? sync_page+0x0/0x50
 [<ffffffff814db743>] io_schedule+0x73/0xc0
 [<ffffffff8110d40d>] sync_page+0x3d/0x50
 [<ffffffff814dbfaf>] __wait_on_bit+0x5f/0x90
 [<ffffffff8110d5c3>] wait_on_page_bit+0x73/0x80
 [<ffffffff8108e1c0>] ? wake_bit_function+0x0/0x50
 [<ffffffff811232d5>] ? pagevec_lookup_tag+0x25/0x40
 [<ffffffff8110d9db>] wait_on_page_writeback_range+0xfb/0x190
 [<ffffffff8110dba8>] filemap_write_and_wait_range+0x78/0x90
 [<ffffffff811a0abe>] vfs_fsync_range+0x7e/0xe0
 [<ffffffff811a0b8d>] vfs_fsync+0x1d/0x20
 [<ffffffff811a0bce>] do_fsync+0x3e/0x60
 [<ffffffff811a0c20>] sys_fsync+0x10/0x20
 [<ffffffff8100b172>] system_call_fastpath+0x16/0x1b
INFO: task iozone:2156 blocked for more than 120 seconds.
"echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
iozone        D 0000000000000001     0  2156   2136 0x00000080
 ffff88003a9abcc8 0000000000000086 ffff88003a9abc90 ffff88003a9abc8c
 ffff88003a9abc38 ffff88003f823240 ffff880002215f80 0000000100129308
 ffff8800374ec6b8 ffff88003a9abfd8 000000000000f598 ffff8800374ec6b8
Call Trace:
 [<ffffffff8110d3d0>] ? sync_page+0x0/0x50
 [<ffffffff814db743>] io_schedule+0x73/0xc0
 [<ffffffff8110d40d>] sync_page+0x3d/0x50
 [<ffffffff814dbfaf>] __wait_on_bit+0x5f/0x90
 [<ffffffff8110d5c3>] wait_on_page_bit+0x73/0x80
 [<ffffffff8108e1c0>] ? wake_bit_function+0x0/0x50
 [<ffffffff811232d5>] ? pagevec_lookup_tag+0x25/0x40
 [<ffffffff8110d9db>] wait_on_page_writeback_range+0xfb/0x190
 [<ffffffff8110dba8>] filemap_write_and_wait_range+0x78/0x90
 [<ffffffff811a0abe>] vfs_fsync_range+0x7e/0xe0
 [<ffffffff811a0b8d>] vfs_fsync+0x1d/0x20
 [<ffffffff811a0bce>] do_fsync+0x3e/0x60
 [<ffffffff811a0c20>] sys_fsync+0x10/0x20
 [<ffffffff8100b172>] system_call_fastpath+0x16/0x1b
Comment 1 Saurabh 2011-11-03 11:16:58 UTC 
for me iozone -a over nfs mount for distribute-replicate is a pass, though it had taken lot of to finish.
Comment 2 Saurabh 2011-11-03 11:17:33 UTC 
(In reply to comment #1)
> for me iozone -a over nfs mount for distribute-replicate is a pass, though it
> had taken lot of time to finish.
Comment 3 Anush Shetty 2011-11-09 06:56:08 UTC 
Observation 3.2.5qa6 - Increasing thread-count helped and is faster.
Comment 4 shishir gowda 2011-11-10 05:43:18 UTC 
Iozone ran to completion for me on a dist-rep setup.
Note You need to log in before you can comment on or make changes to this bug.