Hi ,

I have 100 HP blade servers installed with Suse 10.1 (64 bit ) with Maya 8.5 ( 64 bit ) running on it.

Off late the servers are found to be hung up and cease to respond. The display is also not available. The hung server can be pinged successfully though.

Where exactly can I look for the reasons ? What log files can I refer for the same ?
Where can I check the memory status when the server got hung up ?

Thanks in advance.

If you can ping, can you also ssh into one of the servers? Check the kernel's log, /var/log/messages and any other that may give a clue. I'm thinking that a problem with a device or kernel module may be the last thing logged if you are lucky. You may want to log memory usage on one of the blades before a lock up. Perhaps a program has a memory leak. Check that one of the partitions isn't 100% full.

Utilities like top and vmstat may help. Some have options that allow logging count number of times every n seconds. Monitor the resources periodically and see if you have a gradual increase. There is a book put out by HP "Optimizing Linux Performance" that goes into great detail how to use many performance related tools.

Does the freeze up happen when you aren't running Maya?