[Doublearon]

Here at my new job, we've got a small network hacked together with a smattering of PowerMac G4s, two eMacs, some Linksys routers, 2 networked laser printers, 2 Windows boxes, and an Xserve, all running through a Macsense router into a cable modem.

The problem is, all of the G4s (eMacs included) seem to lock up completely for 30 seconds or so, several times an hour. They're all running OS 9 because of a claimed incompatibility with one of their software programs and OS X. The mouse still moves around, but everything else is locked, and the arrows in the upper left corner are solid.

Notes:
1) the printers aren't the source of the problem... one of them wasn't here when the problem started, and the other one has been disconnected before, and the problem persisted.

2) during the lockup, the Xserve (which is running OS X Server 10.2.6) freezes as well, and shows a spike in usage (the blue lights on the front bezel) right at the end of the freeze.

Does anyone have any ideas I could investigate? I'd really appreciate some direction.

[Doublearon]

update:

1) the freezes are regular, once every 15 minutes
2) I shutdown the Xserve to try and localize the problem, and all problems ceased
3) with the Xserve running, I watched the system processes for something odd, and I came across this: Screenshot 1 Screenshot 2

All I've got to say is WTF ?!

[Camelot]

Originally posted by Doublearon:
update:

1) the freezes are regular, once every 15 minutes
2) I shutdown the Xserve to try and localize the problem, and all problems ceased
3) with the Xserve running, I watched the system processes for something odd, and I came across this: Screenshot 1 Screenshot 2

All I've got to say is WTF ?!

Run top in a terminal window when this happens. You'll get much more info than Process Viewer.app will tell you, including the PID of the process in question.

[Doublearon]

Originally posted by Camelot:
Run top in a terminal window when this happens. You'll get much more info than Process Viewer.app will tell you, including the PID of the process in question.

well, here's the wrinkle... the system freezes (as does the rest of the network). it makes it kind of hard to keep track of usage when the system is suddenly useless. or maybe I should SSH in from an unaffected machine and keep track?

that still doesn't quite explain the errant processes... any ideas?

[Doublearon]

ran tail on the system log... turned out the bonehead who set up the system had NetInfo running to some unknown server (non-existant IP) on the network. so every 15 minutes there was a NetInfo error and an RPC timeout.

yay.

[ghporter]

It's only really "yay" if somebody gets to kick the bonehead-who either entered an invalid IP to start with, or changed a server's IP without checking who may have need for it. Either way, he needs a size 10 up the wazoo.

Wow, that rant made me feel much better!

[Doublearon]

Originally posted by GHPorter:
It's only really "yay" if somebody gets to kick the bonehead-who either entered an invalid IP to start with, or changed a server's IP without checking who may have need for it. Either way, he needs a size 10 up the wazoo.

Wow, that rant made me feel much better!

Apparently, so I've learned, he's an employee's ex-husband. She's got the printout in hand to shove in his face.

[ghporter]

I feel even better now! Thanks for the follow up!