Slow memory leak?

I have a RedHat 7.2 based server that seems to have a slow memory leak
somewhere.

I first noticed it last week - performance went down the tubes, response
time went up, so I checked it - it was swapping like mad. OK, so I added
swap (this particular box has 512 MB RAM, and is hardly ever stressed, so
it was running without swap - quite well - for the last several years.)
It's just that we have not lost power in quite a while, and now something
has eaten all the memory.

I added swap, and it immediately swapped out 68 MB; overnight this has
crept up to 74 MB. At this pace, I'll run out of swap in a few weeks.

[root@pythia root]# free
total used free shared buffers cached
Mem: 513568 509024 4544 0 1932 63172
-/+ buffers/cache: 443920 69648
Swap: 273088 74592 198496
[root@pythia root]# uptime
4:30am up 147 days, 20:05, 4 users, load average: 0.42, 0.32, 0.28
[root@pythia root]# uname -a
Linux pythia.llc.lan 2.4.17-1 #12 SMP Fri Mar 15 06:37:36 EST 2002 i686 unknown

Looking at top and ps output, I don't have any processes running showing
up that are consuming large amounts of memory.

I've restarted all of the basic things that I can think of that might have
a memory leak, but no joy.

I'm looking for a more systematic approach to figuring out what's going on
- how do I track down what's using memory, and what might be leaking it?

TIA,

-Dondo

--
o__
,>/'_ o__
(_)\(_) ,>/'_ o__
What am I on? (_)\(_) ,>/'_ o__
I'm on my bike, (_)\(_) ,>/'_ o__
6 hours a day, busting my ass. (_)\(_) ,>/'_
What are you on? --Lance Armstrong (_)\(_)