[Bug 198313] UFS write becomes very slow in some cases
bugzilla-noreply at freebsd.org
bugzilla-noreply at freebsd.org
Thu Mar 5 13:46:57 UTC 2015
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=198313
Bug ID: 198313
Summary: UFS write becomes very slow in some cases
Product: Base System
Version: 10.1-STABLE
Hardware: Any
OS: Any
Status: New
Severity: Affects Many People
Priority: ---
Component: kern
Assignee: freebsd-bugs at FreeBSD.org
Reporter: demon at FreeBSD.org
CC: mckusick at FreeBSD.org
Hello,
I have 94TB h/w raid6 volume with UFS filesystem on it:
tunefs: POSIX.1e ACLs: (-a) disabled
tunefs: NFSv4 ACLs: (-N) disabled
tunefs: MAC multilabel: (-l) disabled
tunefs: soft updates: (-n) enabled
tunefs: soft update journaling: (-j) disabled
tunefs: gjournal: (-J) disabled
tunefs: trim: (-t) disabled
tunefs: maximum blocks per file in a cylinder group: (-e) 4096
tunefs: average file size: (-f) 16384
tunefs: average number of files in a directory: (-s) 64
tunefs: minimum percentage of free space: (-m) 8%
tunefs: space to hold for metadata blocks: (-k) 9136
tunefs: optimization preference: (-o) time
tunefs: volume label: (-L)
When it is about 5TB free on it (which does not count 8% of reserved space)
/dev/mfid0p1 94T 82T 5.1T 94% /opt
all write operations on it become very-very slow (1-2 MB/sec). The process
which writes data consumes 100% of CPU in system state:
{agata}/home/mitya% ps -p 58796 -o systime,usertime
SYSTIME USERTIME
0:59.56 0:00.08
58796 mitya 1 103 0 27808K 4820K CPU11 11 0:53 100.00% bsdtar
As soon as I free some space, the problem disappears until space fills up
again.
Since I have the recommended 8% of reserved free space, I expect writes to work
well until all free space is exhausted.
--
You are receiving this mail because:
You are the assignee for the bug.
More information about the freebsd-bugs
mailing list