NFS Services Administrator's Guide

Troubleshooting NFS Services
Performance Tuning
Chapter 8334
To Adjust the Number of nfsd Processes
1. Issue the following command on the NFS server:
netstat -s
If the UDP statistics displayed by the netstat command indicate a
large number of socket overflows, as in the following example, then
your server is not running enough nfsd daemons.
udp:
0 incomplete headers
0 bad data length fields
0 bad checksums
1375 socket overflows
2. To increase the number of nfsd daemons that will start at the next
system boot (or when restarting the NFS services by running the
/sbin/init.d/nfs.server start command), change the value of
the NUM_NFSD variable in the /etc/rc.config.d/nfsconf file, as in
the following example:
NUM_NFSD=64
3. Issue the following command to directly start more nfsd processes
once you ensure there is no nfsd process currently running on the
system:
/usr/sbin/nfsd num_nfsd
The value of the num_nfsd option is the suggested number of nfsd
daemons that will start. The actual number of daemons started is
one daemon to support kernel TCP threads, plus a number of UDP
daemons. The number of UDP daemons started is the value of
num_nfsd rounded up to a multiple of the number of active CPUs in
the system.
For more information, type man 1M nfsd at the HP-UX prompt.
4. Issue the netstat -s command again to check the number of socket
overflows. Continue to adjust the NUM_NFSD value and start nfsd
processes until the number of new socket overflows is close to zero.
(The output of nfsstat is cumulative, so when there are no new
socket overflows, the number will stay the same.)