public inbox for linux-xfs@vger.kernel.org
 help / color / mirror / Atom feed
* XFS Syncd
@ 2015-04-10  4:23 Shrinand Javadekar
  2015-04-10  6:32 ` Dave Chinner
  0 siblings, 1 reply; 21+ messages in thread
From: Shrinand Javadekar @ 2015-04-10  4:23 UTC (permalink / raw)
  To: xfs

[-- Attachment #1: Type: text/plain, Size: 1499 bytes --]

Hi,

I am using the XFS filesystem as the backend for Openstack Swift. On
my setup, I have a single server with 8 data disks; each of them is
one XFS volume.

I am running a workload which does many concurrent writes of 256K
files into the XFS volumes. Openstack Swift takes care of evenly
distributing the data across all the 8 disks. It also uses extended
attributes for each of the files it writes. It also explicitly does a
fsync() at the end for each file.

I am seeing a behavior where the system pretty much stalls for ~5
seconds after every 30 seconds. I see that the # of ios goes up but
the actual write bandwidth during this 5 second period is very low
(see attached images). After a fair bit of investigation, we've
narrowed down the problem to XFS's syncd (fs.xfs.xfssyncd_centisecs).
This runs at a default interval of 30 seconds.

I have a couple of questions:

1. If all file writes are done with an fsync() at the end, what is
xfssyncd doing for several seconds?
2. How does xfssyncd actually work across several disks? Currently, it
seems that when it runs, it pretty much stalls the entire system.
3. I see that fs.xfs.xfssyncd_centisecs is the parameter to tune the
interval. But that doesn't give us much. Increasing the interval
simply postpones the work. When xfssyncd runs, it takes more time. Are
there any other options I can try to make xfssyncd not stall the
system when it runs?

Thanks in advance.
-Shri

P.S. I'm not a member of this list. Direct replies appreciated.

[-- Attachment #2: write_throughput6.png --]
[-- Type: image/png, Size: 109367 bytes --]

[-- Attachment #3: read_write_requests_complete_rate6.png --]
[-- Type: image/png, Size: 93756 bytes --]

[-- Attachment #4: Type: text/plain, Size: 121 bytes --]

_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs

^ permalink raw reply	[flat|nested] 21+ messages in thread

end of thread, other threads:[~2015-06-09 23:13 UTC | newest]

Thread overview: 21+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2015-04-10  4:23 XFS Syncd Shrinand Javadekar
2015-04-10  6:32 ` Dave Chinner
2015-04-10  6:51   ` Shrinand Javadekar
2015-04-10  7:21     ` Dave Chinner
2015-04-10  7:29       ` Shrinand Javadekar
2015-04-10 13:12         ` Dave Chinner
2015-06-02 18:43           ` Shrinand Javadekar
2015-06-03  3:57             ` Dave Chinner
2015-06-03 23:18               ` Shrinand Javadekar
2015-06-04  0:35                 ` Dave Chinner
2015-06-04  0:58                   ` Shrinand Javadekar
2015-06-04  1:55                     ` Dave Chinner
2015-06-04  1:25                   ` Dave Chinner
2015-06-04  2:03                     ` Dave Chinner
2015-06-04  6:23                       ` Dave Chinner
2015-06-04  7:26                         ` Shrinand Javadekar
2015-06-04 22:08                           ` Dave Chinner
2015-06-05  0:59                         ` Shrinand Javadekar
2015-06-05 17:31                           ` Shrinand Javadekar
2015-06-08 21:56                             ` Shrinand Javadekar
2015-06-09 23:12                               ` Dave Chinner

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox