1
Technical Help: Problems BEFORE entering the game /
« on: July 22, 2005, 11:29:13 pm »
The server was down due to a faulty CPU fan. This caused the machine to shut down automatically 3 times in the past week.
On the first occasion we were notified about 1 day after the server stopped responding. We do not actively keep tabs on the server but I check my e-mail very frequently and we did not hear anything until about 24 hours after the shutdown.
The second time the server went down was about 3 days after the first instance. We sent in a restart request a few hours after I noticed the server went down.
However, the server went down again 1 hour after the 2nd reboot. I went to the data centre a few hours later to look at what went wrong. The problem was identified soon after and I swapped out the faulty CPU fan.
Granted we ought to keep tabs on the server, some of the comments here make it seem that our server shuts itself down on a weekly basis. You\'d notice that our uptime is actually pretty decent based on statistics at http://gandalf.fragnetics.com/status/system/proc.html
We had intended to swap out the CPU fan sometime last year but decided against it then. This problem is likely to surface again in about 1 year but I think we know what to do should something similar happen again. The fan gets spoilt more easily when the server is heavily utilized, and the recent increase in CPU usage by the PS server accelerated this process. My bad for not realizing it might have been due to this problem earlier - we have some experience with faulty CPU fans.
On the first occasion we were notified about 1 day after the server stopped responding. We do not actively keep tabs on the server but I check my e-mail very frequently and we did not hear anything until about 24 hours after the shutdown.
The second time the server went down was about 3 days after the first instance. We sent in a restart request a few hours after I noticed the server went down.
However, the server went down again 1 hour after the 2nd reboot. I went to the data centre a few hours later to look at what went wrong. The problem was identified soon after and I swapped out the faulty CPU fan.
Granted we ought to keep tabs on the server, some of the comments here make it seem that our server shuts itself down on a weekly basis. You\'d notice that our uptime is actually pretty decent based on statistics at http://gandalf.fragnetics.com/status/system/proc.html
We had intended to swap out the CPU fan sometime last year but decided against it then. This problem is likely to surface again in about 1 year but I think we know what to do should something similar happen again. The fan gets spoilt more easily when the server is heavily utilized, and the recent increase in CPU usage by the PS server accelerated this process. My bad for not realizing it might have been due to this problem earlier - we have some experience with faulty CPU fans.