Network and Storage Protocols

Any Way To Check Running Processes?

TOM_SMITH7444
11,460 Views

I've been looking online for a method to check what is currently running on our FAS2240 controllers, but so far have come up blank.

I can get plenty of stats using commands like "sysstat -u 1" - which is what we use to get a general view of how "busy" the SAN is. But when the CPU gets a little high, for example, I'd like to be able to determine what's causing it. Usually I check things like sis, snapmirror, volume creation / deletion etc, but this doesn't always indicate where the high usage lies.

Anyone have any ideas on how to check what specific processes are running, and ideally how much resources each is using? I'm basically after a "ps -ef" type command.

Thanks

Tom

1 ACCEPTED SOLUTION

DOUGLASSIGGINS
11,329 Views

ps is a command from the cli:

Typically I run something like ps -c 4

usage: ps [-l] <pid> ...

       ps [-hlps] [-c cutoff]

       ps -z

          -c cutoff  limit output to threads above the cutoff point

          -h         equivalent to -c 5

          -l          long listing

          -p          skip threads which don't have profile information

          -s          use scheduling time instead of total time

          -z          clear the times/counts in the proc structure

Another handy command is:

sysstat -M 1

Or statit.

What you are probably seeing is the ANY cpu get up above 90% These three commands should help you zero in on the issue

View solution in original post

4 REPLIES 4

HENRYPAN2
11,328 Views

Tom,

You may wish to try Balance.

Good luck + Good Friday

Henry

TOM_SMITH7444
11,328 Views

Thanks Henry. I'll look into Balance.

DOUGLASSIGGINS
11,330 Views

ps is a command from the cli:

Typically I run something like ps -c 4

usage: ps [-l] <pid> ...

       ps [-hlps] [-c cutoff]

       ps -z

          -c cutoff  limit output to threads above the cutoff point

          -h         equivalent to -c 5

          -l          long listing

          -p          skip threads which don't have profile information

          -s          use scheduling time instead of total time

          -z          clear the times/counts in the proc structure

Another handy command is:

sysstat -M 1

Or statit.

What you are probably seeing is the ANY cpu get up above 90% These three commands should help you zero in on the issue

TOM_SMITH7444
11,327 Views

Thanks Doug - I tried ps from the "normal" CLI but didn't think to change to advanced mode.

Cheers

Public