Bug 187414 - [zfs] ZFS Write Deadlock on 8.4
Summary: [zfs] ZFS Write Deadlock on 8.4
Status: Closed Overcome By Events
Alias: None
Product: Base System
Classification: Unclassified
Component: kern (show other bugs)
Version: 8.4-STABLE
Hardware: Any Any
: Normal Affects Only Me
Assignee: freebsd-fs (Nobody)
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2014-03-10 11:40 UTC by Kuo-Cheng Yeu
Modified: 2018-05-29 10:02 UTC (History)
0 users

See Also:


Attachments

Note You need to log in before you can comment on or make changes to this bug.
Description Kuo-Cheng Yeu 2014-03-10 11:40:00 UTC
Just did a periodical update to 8.4-STABLE r262357 on a self-build home
server. The only change I did after update is upgrade zpool version to
latest (version 28). My zpool is in mirror configuration, built from dual
Seagate 3TB hard drives (ST3000DM001). 


Moving large files in ~4GB size hangs the target directory, and Bonnie++
leads to deadlock as well. So I believe it was caused by large write
operations. 

Here are logs for both cases. 
mv: procstat -kk -a 
http://pastebin.com/WManZzbv 

bonnie++: procstat -kk -a 
http://pastebin.com/fe2bmbbY 

bonnie++: kgdb (thread apply all bt) 
http://pastebin.com/kdm5xQwM

How-To-Repeat: run bonnie++ and set -d scratch-dir on ZFS.
Comment 1 Kuo-Cheng Yeu 2014-03-18 09:11:51 UTC
Clean install FreeBSD 10.0-STABLE r263205, everything works fine. Time to
say goodbye to 8.4-STABLE.

uname -a
FreeBSD kmd.twbbs.org 10.0-STABLE FreeBSD 10.0-STABLE #0 r263205: Sat Mar
15 23:38:52 CST 2014     root@kmd.twbbs.org:/usr/obj/usr/src/sys/MERCURY
 amd64
Comment 2 Mark Linimon freebsd_committer freebsd_triage 2014-04-16 02:43:20 UTC
State Changed
From-To: open->suspended

Clean up synopsis and assignment.  Setting to Suspended since submitter 
notes it no longer happens with 10.0, but leaving in GNATS to warn others. 


Comment 3 Mark Linimon freebsd_committer freebsd_triage 2014-04-16 02:43:20 UTC
Responsible Changed
From-To: freebsd-bugs->freebsd-fs
Comment 4 john 2014-04-16 14:39:26 UTC
-- 
John Baldwin
Comment 5 John Baldwin freebsd_committer freebsd_triage 2014-04-16 14:40:31 UTC
If the processes are hung in the "dmu_tx_delay" wait channel, then this might 
have been fixed by:

http://svnweb.freebsd.org/base?view=revision&revision=264505

-- 
John Baldwin
Comment 6 Eitan Adler freebsd_committer freebsd_triage 2018-05-28 19:50:09 UTC
batch change:

For bugs that match the following
-  Status Is In progress 
AND
- Untouched since 2018-01-01.
AND
- Affects Base System OR Documentation

DO:

Reset to open status.


Note:
I did a quick pass but if you are getting this email it might be worthwhile to double check to see if this bug ought to be closed.
Comment 7 Andriy Gapon freebsd_committer freebsd_triage 2018-05-29 10:02:36 UTC
Either fixed or not reproducible.