As you all have noticed, the server has been plagued by stability problems for the past few days. For the record, this is the first time that the site'* availability problems have been directly related to the server itself. I have run out of ideas for how to fix the problem, so I'm looking for outside suggestions.

Essentially, every day there has been a problem, the server stops communicating with the rest of the network at about 2:00 p.m. CDT , +/- 2 hours. At 1:30 a.m., +/- 1 hour, communications resumes. During the intervening time, the server seems to stay happily active, normal in all aspects except for its isolation from the world. None of the logs signify anything out of the ordinary, and my IDS shows no indication of foul play. A restart seemed to cure things for a little while, but when the problem resurfaced, an additional restart did not alleviate the problem. There are no cron jobs that run near the specified times. Initially I suspected hardware failure, but the presence of the clean log files disputes that theory. During the problem hours, none of the computers on the server'* LAN can ping the server.

I have no idea what might be causing the server to do such things. If any of you can help me, please e-mail me at teuobk@ringworld.org (on a different server, at a different site, so it shouldn't be affected). I would really, really appreciate any and all ideas.

Thanks,
Jeff

The server didn't go deaf today, so hopefully the problem is fixed (for real this time!)

If true, then the solution was to add a static entry for the gateway into the server'* ARP table.

The problem resurfaced today. Deaf at 4:30 p.m., recovery at 1:30 a.m.

I'm baffled, so please please please, if you have any ideas, inklings, psychic powers, ANY SUGGESTIONS AT ALL, please let me know!

Thanks!