*** Reply to note of Fri, 22 Nov 2002 02:42:13 -0500 (EST/CDT)
*** by [EMAIL PROTECTED]

You could try to increase the number of server threads for the nfs server
in /etc/rc.config:

   #
   # the kernel nfs-server supports multiple server threads
   #
   USE_KERNEL_NFSD_NUMBER="16"

You can also use the nfsstat command (nfsstat -s for server and
nfsstat -c for the client) to gather more information.

sal

Michael MacIsaac <[EMAIL PROTECTED]> writes:
>Hi list,
>
>We're seeing network problems (especiallly NFS) under heavy loads.  The
>configuration is G6 with OSA Express Gigabit Ethernet.  Linux images are
>SLES-7 with patched (.8) qdio/qeth drivers sharing the OSA. In the NFS
>server's /var/log/messages we see constant:
>
>Nov 20 15:16:30 linuxnfs kernel: ch-0c71: System reset (remote)
>Nov 20 15:16:35 linuxnfs kernel: ch-0c21: System reset (remote)
>Nov 20 15:16:35 linuxnfs kernel: ch-0c41: System reset (remote)
>Nov 20 15:16:35 linuxnfs kernel: ch-0c31: System reset (remote)
>Nov 20 15:16:35 linuxnfs kernel: ch-0c51: System reset (remote)
>Nov 20 15:16:35 linuxnfs kernel: ch-0c61: System reset (remote)
>Nov 20 15:16:35 linuxnfs kernel: ch-0c81: System reset (remote)
>Nov 20 15:16:35 linuxnfs kernel: ch-0c11: System reset (remote)
>Nov 20 15:16:35 linuxnfs kernel: ch-0c71: System reset (remote)
>Nov 20 15:16:40 linuxnfs kernel: ch-0c21: System reset (remote)
>
>Any ideas? TIA
>
>          -Mike MacIsaac,  IBM   [EMAIL PROTECTED]   (845) 433-7061

Reply via email to