On Sun, 8 May 2011 at 18:03 -0000, Dave Love wrote:
> >> Is it worth worrying about any real accuracy in accounting?
> >> Tyring to keep track of everything relevant, like system problems
> >> that clobber jobs, just doesn't seem worthwhile to me.
>
> > This is a very important basic question and strongly it interacts
> > with the accuracy of the collected information.
>
> I don't think it has much to do with what GE reports, but it seems I
> don't understand.
It is a question the user needs to ask himself. "Is the available SGE
reporting sufficient for my needs? Do I even understand what SGE
reports?"
I may also be over thinking the problem. Several times in the past
I've worked on systems which collected data that was later turned in
to bills which caused real money to change hands.
> I'm surprised the resource manager would be expected to keep track
> of resources it doesn't manage,
Correct. If there is other use it might need to be feed into the
accounting data feed if it needs to be tracked.
> and that you run stuff on the nodes outside it -- that sounds like
> asking for trouble.
Not intentionally. We have had users ssh into random compute nodes
and run things (not maliciously). Getting the appropriate pam modules
in place is also on our long TODO list.
I've read of MPI jobs escaping out of the SGE process control and
various things to deal with this in various ways.
If you are only interested in basic usage reporting, qacct is probably
fine. It sound like this applies to you, I think it applies to us and
it probably applies to most other SGE operators.
> > We will probably also need to do a web page showing some sort of
> > historical reporting. This will need to integrate into our web
> > content management system.
> Do share anything you can that's sufficiently general to be useful
> others directly or as examples, of course.
I do hope to do so if we get anything useful. It will probably have
site specifics and need adaption for someone else.
Stuart
--
I've never been lost; I was once bewildered for three days, but never lost!
-- Daniel Boone
_______________________________________________
users mailing list
[email protected]
https://gridengine.org/mailman/listinfo/users
