Q: My system seems to run out of swap (paging) space. The swapfile,
/private/vm/swapfile, grows and grows and grows very large. If I
check the process status of the process running lookupd, I notice
that it has grown very large. What's going on?
A: ANYONE EXPERIENCING THIS PROBLEM WITH RELEASE 1.0 SHOULD UPGRADE
TO RELEASE 1.0a. THIS IS A BUG IN 1.0 THAT IS FIXED IN 1.0a.
The bug in the Releae 1.0 lookupd is only tickled when NIS (formerly known as Yellow Pages or "YP") is in use. It's a memory leak; it is exercised especially
when accessing the 'group' database (i.e., /etc/group). There are
other leaks as well; group is the worst of them.
The most likely symptom of this is that your disk might fill
up. This is because the swapfile (/private/vm/swapfile) grows very
large, because lookupd grows very large. At one site, where they have
600+ nodes on their (single) network, with many, many different
manufacturers' machines, all running YP, lookupd was seen to grow 16MB
overnight, with the only things running being those in the standard
daemon set, and those triggered by the standard crontab.
