Here is my question:
OS: SuSE 6.2 kernel 2.2.5 (In servers and client)
I have a PC running like a nfs-client (called "concentrador") with an
application that have to poll 3 PC's nfs-servers (called "puesto1",
"puesto2", "puesto3") in a continuous loop.
When a nfs-server (for example "puesto2") make a request to the nfs-client
("contrador"), the answer of "concentrador" is so slow (more than 5 o2 10
seconds) that "puesto2" suposes that "concentrador" is down.
Making the same with a PC nfs-client and 2 PC's nfs-servers performance is
OK and there's no problem at all.
The way server and client can communicate is by writing in some files
residents in nfs-server file system (in the exported directory).
I've modified the number of nfs daemons (up to 12 in a PC) (on the client
side and ond the server side).
I've mounted the nfs-servers with differents options:
actimeo=1
actimeo=0
rsize=4096, wsize=4096
But none of this worked.
Looking into /var/log/messages of the nfs-client PC I've found this:
nfs: RPC call returned error 111
RPC: task of released request still queued
RPC: (task is on xprt_pending)
and also this:
kernel: __nfs_fhget: inode 92320 busy, icount=2, inlink=1
kernel: __nfs_free_dentries: found //NEGATIVA.cag, d_count=0, hashed=1
kernel: nfs_dentry_delete: //NEGATIVA.cab; ino=92320, count=2, i_nlink=1
Lots of thanks in advance for your help !
Enrique Bernal
ebv@a68.com
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/