Hi,

I have run prstat -t 1 on a Solaris 9 server which has been under pressure recently. This is the output I have got.

NPROC USERNAME SIZE RSS MEMORY TIME CPU
403 oracle 804G 705G 100% 33:34:18 62%

Here is the explanation for SIZE and RSS.

SIZE The total virtual memory size of the process, includ-
ing all mapped files and devices, in kilobytes (K),
megabytes (M), or gigabytes (G). The resident set size
of the process (RSS), in kilobytes (K), megabytes (M),
or gigabytes (G).


How can the size be 804GB? The CPU is also regularly up to 100% for a period of time.

Total memory physical RAM in the server is 8GB.

Memory size: 8192 Megabytes

Total swap:

# swap -s
total: 2757296k bytes allocated + 683680k reserved = 3440976k used, 5427800k available

Thanks.

Check your swap space. Our server does that when it runs out of swap space. Also check your temp (tmp) directory to make sure it isn't to full. Mail could also be another culprit. Just a few things I have run across.

Check your disk i/o during that surge and also the swap might be running out

that's because solaris counts the amount of memory per process, so if you run your oracle db there are many processes that uses the same memory area, that's called shared-memory, e.g your oracle SGA... for example, if you have 5 processes that use 2 GB as shared memory, solaris will say, that 10GB of memory are in use, based on the sum... look at your ipcs -a output

if your sun is swapping you have to less memory, you could add more DIMMs or tune your database... and if the load is around 90%, your mashine is not powerfull enough to run your applications...

gP

Thanks for your help everyone. I just wanted to confirm the obvious. The machine is not powerful enough to run the applications. Thanks.