I have a solaris 10 fileserver, with some NFS exports.  

It is shared by many compute servers in an SGE cluster.

To alleviate the bottleneck, all the compute servers have local disk for
scratch space, that users are supposed to use for any IO intensive workload.

 

Often times, some user neglects to configure their job correctly, and they
end up hammering the fileserver.  And then people complain of slowness, and
I have to figure out what's wrong ...  Which usually means finding out who
is hammering the server, from what machine, using what process.

 

I'd like to improve my abilities to find the cause of these problems.  So
here's my question:

 

Is there some way, on the server, for me to identify

                Most of the NFS traffic is coming from host A and B

                or

                There's a lot of activity in the file or dir
/exports/foo/bar/baz

 

The answers to these questions would greatly improve my ability to identify
who's hammering the server.  Thank you...

_______________________________________________
opensolaris-help mailing list
opensolaris-help@opensolaris.org

Reply via email to