[RESOLVED] GFS2 performance issues

[npavlidis]

Hello all,

I have setup a gf2 cluster following this:
[url=http://www.linuxdynasty.org/howto-setup-gfs2-with-clustering.html]how-to-gfs2-cluster[/url]

I have followed this:
[url=http://www.linuxdynasty.org/howto-increase-gfs2-performance-in-a-cluster.html]performance-howto[/url]
and
[url=http://docs.redhat.com/docs/en-US/Red_Hat_Enterprise_Linux/6/html/Global_File_System_2/gfs2_performance_troubleshoot.html]red-hat-performace-troubleshoot[/url]

debugfs is mounted on www1 at the moment but I cant make much of it


[root@www1 ~]# dd if=/dev/urandom of=/webs/test.dat bs=104857600 count=1
1+0 records in
1+0 records out
104857600 bytes (105 MB) copied, 34.4229 seconds, [b]3.0 MB/s [/b]

root@www1 ~]# ./ping_pong /webs/test.dat 4
707 locks/sec

[root@www1 ~]# ./ping_pong -rw /webs/test.dat 4
data increment = 1
713 locks/sec

these values keep going up and down (ping_pong ones not the dd one)
(up and down ranged from 1400 to 30 and 300)

NOTE: Even though I am using CentOS 5.6 I am opening the thread here since I am using a later kernel (2.6.32.43-6.art.x86_64) in any case.. please let me know if you would like me to open the thread on the 5.x section

Any help would be greatly appreciated

Thank you in advance,

Nik

[npavlidis]

maybe I shouldn't have been using /dev/urandom (how stupid do I feel!!)
sorry
my bad.

[root@www1 httpd]# dd if=/dev/zero of=/webs/test.dat bs=104857600 count=1
1+0 records in
1+0 records out
104857600 bytes (105 MB) copied, 0.86853 seconds, 121 MB/s


no issue here...

[AlanBartlett]

Thank you for reporting back. If you are sure that everything is now correct then I shall, on your behalf and for posterity, mark this thread as [RESOLVED].

[paradigm]

Hi npavlidis ,

did this :

http://www.linuxdynasty.org/howto-increase-gfs2-performance-in-a-cluster.html

make a big diffrence in preformance ? and how about stabilty ?