NFS Services Administrator's Guide

Troubleshooting NFS Services
Performance Tuning
Chapter 8338
To Improve NIS+ Performance
❏ Issue the following command to check the size of your transaction
log:
/usr/lib/nis/nislog | head -10
If your transaction log is fully checkpointed, it will contain only three
entries. If it contains many entries, issue the following command to
checkpoint it:
nisping -Ca
❏ The nisping -C command can cause a long delay if your namespace
is large. Do not reboot the system. Do not reenter the nisping
command. This problem will solve itself. Just wait until the server
finishes checkpointing.
❏ Make sure your NIS_PATH environment variable is set to something
clean and simple, like org_dir.$:$. A complex NIS_PATH value,
particularly one that contains a variable, will slow your system and
may cause some operations to fail. See “To Change the Search Order
of Domains” on page 237.
❏ Concatenation paths in tables slow performance. If performance is a
problem in your NIS+ namespace, do not use concatenation paths.
See “To Create or Remove Paths Among Tables” on page 256.
❏ Make sure you have 10 or fewer replica servers per domain.
❏ NIS+ groups that contain other groups (recursive groups) slow NIS+
performance. If performance is a problem in your NIS+ namespace,
do not use recursive NIS+ groups. See “To Add or Remove Members
of an NIS+ Group” on page 258.
❏ Large transaction logs slow NIS+ performance, particularly at
system startup. If your transaction logs are large, or if you have just
run the nispopulate script to populate your domain tables, issue the
nisping -Ca command to checkpoint your directories. Make sure
your master server has a cron job scheduled to issue the nisping
-Ca command daily. Type man 1 crontab for information.
❏ Issue the ps -ef command to make sure nis_cachemgr is running
on every NIS+ client host. Start it if it is not. Type man 1M
nis_cachemgr for information.
❏ An NIS+ lookup command like niscat returns the error message