Hi,
I wanted to make sure people on this thread knew that we are about to
start on a gram audit v2 campaign. Discussions will take place on the
gram-dev list. http://www.globus.org/mail_archive/gram-dev/2008/04/msg00003.html
Join the gram-dev list of you want to follow along/participate. http://dev.globus.org/wiki/GRAM#Mailing_Lists
There will be a telecon tomorrow Thursday to discuss details at 11 am
central.
Thanks,
Stu
On Feb 13, 2008, at Feb 13, 2:44 PM, Markus Binsteiner wrote:
Hi.
I'm working on the ARCS project (former APACGrid) in Australia and
we also would like to see logging of some of the information as
described in the original message.
In addition, it would be useful if the logging database would
contain the
time when the job actually started on the cluster. I don't know how
feasable that is (although I think to remember that former versions
of globus had a feature like this) but it would be really useful to
display some kind of percentage of how much of the actual walltime
the job already went through in submission clients.
Also, it would be nice if the audit database could store the FQAN of
the
VO that is used to submit the job. Since there is not a one-to-one
relationship of VO <-> local user at the moment it's not possible
to get this information...
Cheers,
Markus
----- "Frank Breitling" <[EMAIL PROTECTED]> wrote:
Hello,
In WS GRAM audit logging we are missing some important information.
That is:
o global job id (UUID format), which is returned from the host to the
client in the EPR file
o resource usage information as reported by the UNIX time command
consisting of:
(i) the elapsed real time between invocation and termination
(ii) the user CPU time (the sum of the tms_utime and tms_cutime
values
in a struct tms as returned by times(2))
(iii) the system CPU time (the sum of the tms_stime and tms_cstime
values in a struct tms as returned by times(2))
o end time of the job, in the same format as creation_time
o name of submission client
o name of execution host
(o and maybe also the number of used CPUs)
Moreover, we would like to obtain the audit information in Usage
Record XML format developed by the Usage Record working group of the
open grid forum. ( https://forge.gridforum.org/projects/ur-wg/ ,
https://forge.gridforum.org/sf/docman/do/listDocuments/projects.ur-wg/docman.root
).
The information could be written to an XML file for each job. A
natural name would be "UUID.ur".
We would like to ask for an implementation of these features since we
would like to use audit logging also for monitoring of usage
statistics.
I have filed two feature request on this:
http://bugzilla.globus.org/globus/show_bug.cgi?id=5864
http://bugzilla.globus.org/globus/show_bug.cgi?id=5865
Best regards
Frank
