I've seen these symptoms before on one of our development machines - although one that doesn't normally run PingPlotter 24X7 (but does stay on for weeks sometimes with PingPlotter running pretty often during then).

We have another machine here, though (running XP SP2), that runs PingPlotter constantly - and has been for years (literally) - with occasionally reboots for patching. There are no similar symptoms there.

An old version of PingPlotter (in the 1.x line) had some problems that sound similar - caused by creating a new thread for each new packet sent out. After running for a few days, Windows would start to have problems. We fixed this in a later release by recycling threads. This same type of thing could possibly be happening with another resource type, though (handle, network socket, other?) - we just need to identify what that resource is, and find out if PingPlotter is doing something that is inciting some odd behavior on that front.

What packet type are you using (ICMP, UDP, TCP?), and what trace interval? Are you using a software firewall (ie: ZoneAlarm), and if so, what version? How many instances of PingPlotter do you normally run?

We're definitely interested in looking in to this problem further.

- Pete