Could be related to some of the nfs bugs that have been noted already
with rhel5.  make sure your kernel is 2.6.18-53.1.4

-----Original Message-----
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of Alessandro
Tinivelli
Sent: Tuesday, January 29, 2008 11:09 AM
To: [email protected]
Subject: [rhelv5-list] Problem with NFS on new rhel5 kernels

Hi all,

i've experiencing a problem difficult to solve about NFS performance.

I have a web server cluster (running on RHEL4) where all servers mount a

nfs file server (NetApp F270):

[EMAIL PROTECTED] ~]# cat /proc/mounts | grep nfs
fil01:/vol/vol1/web /www nfs 
rw,v3,rsize=8192,wsize=8192,hard,intr,tcp,nolock,proto=tcp,timeo=4,retra
ns=5,addr=fil01 
0 0

Mounting the file server on a new machine where is running a fresh 
install of RHEL5 (don't care about sizes, i have tried all combination 
possible)

[EMAIL PROTECTED] ~]# cat /proc/mounts | grep nfs
fil01:/vol/vol1/web /www nfs 
rw,vers=3,rsize=32768,wsize=32768,hard,intr,nolock,proto=tcp,timeo=4,ret
rans=5,sec=sys,addr=fil01 
0 0

shows a performance problem: the CPU load and ops/s on the NetApp 
increase dramatically, and reading becomes slow

nfsstat command shows an interesting fact: a very high number of lookups

Client nfs v3:
null         getattr      setattr      lookup       access
0         0% 10366     0% 2         0% 975623   90% 63084

at the moment I have some "nested mounting" (a second file server 
mounted on a directory in the filesystem of the main one) and i'm moving

them in a non "nested" way.

Is this a possibile cause of the problem? What other can i try?

Thank you in advance

-- 
Alessandro Tinivelli
[EMAIL PROTECTED]

_______________________________________________
rhelv5-list mailing list
[email protected]
https://www.redhat.com/mailman/listinfo/rhelv5-list

_______________________________________________
rhelv5-list mailing list
[email protected]
https://www.redhat.com/mailman/listinfo/rhelv5-list

Reply via email to