[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]

Re: [Linux-cluster] GFS Tuning - it's just slow, to slow for production



That should be "-o relatime" and "-o noatime", respectively =)


From: linux-cluster-bounces redhat com [mailto:linux-cluster-bounces redhat com] On Behalf Of Alan A
Sent: dinsdag 16 maart 2010 21:01
To: linux clustering
Subject: Re: [Linux-cluster] GFS Tuning - it's just slow,to slow for production

I rebuild the share as GFS2, and mounted GFS2 instead of GFS.

In /etc/fstab I am mounting with:

/dev/mapper/vg_acct10-lv_acct10 /acct10 gfs2 defaults 1 2

I tried mounting:
/dev/mapper/vg_acct10-lv_acct10 /acct10 gfs2 -o reltime 1 2

and
/dev/mapper/vg_acct10-lv_acct10 /acct10 gfs2  -o notime 1 2

It complained about /etc/fstab not having valid arguments. I also got  very different parameters with gfs2_tool gettune /mountpoint then what I saw in gfs:

[root fenclxmrcati11 bin]# gfs2_tool gettune /acct10
new_files_directio = 0
new_files_jdata = 0
quota_scale = 1.0000   (1, 1)
logd_secs = 1
recoverd_secs = 60
statfs_quantum = 30
stall_secs = 600
quota_cache_secs = 300
quota_simul_sync = 64
statfs_slow = 0
complain_secs = 10
max_readahead = 262144
quota_quantum = 60
quota_warn_period = 10
jindex_refresh_secs = 60
log_flush_secs = 60
incore_log_blocks = 1024



My question - how do I mount the volume via /etc/fstab with -o noatime or -o reltime?
For quota management - Documentation suggests mounting with mount -o quota=account options, is this necessary if I set the quotas with gfs2_quota command?

 





On Thu, Mar 4, 2010 at 11:25 AM, Gordan Bobic <gordan bobich net> wrote:
Are you concurrently accessing same directories from multiple nodes? If so, re-arrange the access pattern so each node is accessing independent subtrees. Otherwise you'll have massive lock churn and very poor latencies.

Gordan

"Alan A" <alan zg gmail com> wrote:

>We are trying to deploy GFS in production, and are experiencing major
>performance issues. What parameters in GFS settune can be changed to
>increase I/O, to better tune performance? Application we run utilizes a lot
>of I/O, please advise.
>
>We experience OK performance when starting, but as things ramp up and we get
>few processes going GFS slows dramatically.
>
>--
>Alan A.
--
Sent from my Android phone with K-9 Mail. Please excuse my brevity.

--
Linux-cluster mailing list
Linux-cluster redhat com
https://www.redhat.com/mailman/listinfo/linux-cluster



--
Alan A.

No virus found in this incoming message.
Checked by AVG - www.avg.com
Version: 9.0.790 / Virus Database: 271.1.1/2750 - Release Date: 03/16/10 08:33:00


[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]