Michel Jouvin wrote:
> You need a more recent version of globus-gma. Which version are you
> using ?
>
> Michel
>
> --On mercredi 1 juillet 2009 11:59 +0200 Carlos Fernández Iglesias
> <[log in to unmask]> wrote:
>
>> Hello,
>>
>> I need to decrease the time between job submitted an done (on the UI),
>> modifying the parameter tick in globus-gma.conf.
>>
>> I've found this list of parameters from the globus-gma.conf:
>>
>> * |*logf*| (srting) --- location of the log file (default is
>> relative to GLOBUS_LOCATION)
>> * |*gridservices*| (string) --- path to the gridservices directory
>> /[never change this]/ (default is relative to GLOBUS_LOCATION)
>> * |*agentpath*| (string) --- path to the directory with agent files
>> /[never change this]/ (default is relative to GLOBUS_LOCATION)
>> * |*groups*| (0 or 1) --- if set, supplementary groups will be
>> applied to the poll process (disabled (0) by default)
>> * |*condorfix*| (0 or 1) --- enables a Condor work-around for not
>> distinguishing VOMS attributes (disabled (0) by default)
>> * |*tout*| (numeric) --- sets a limit in seconds for a single job
>> state poll to finish (30 by default)
>> * |*toutlim*| (numeric) --- sets a limit for a number of consecutive
>> poll timeouts for a given user, after which all remaining jobs for
>> that user will be skipped till the next poll cycle (4 by default)
>> * |*tick*| (numeric) --- number of seconds between poll cycles.This
>> parameter defines granularity for |*stateage*|, |*fileage*| and
>> adaptive state refresh interval below (300 by default).
>> * |*stateage*| (numeric) --- number of seconds for which a job state
>> is considered 'fresh' (600 by default)
>> * |*statefact*| (numeric) --- division factor for calculating
>> adaptive state refresh interval for short jobs /[refresh_interval
>> = min(stateage, job_run_time / statefact)]/ (disabled (0) by
>> default)
>> * |*fileage*| (numeric) --- number of seconds before a job file is
>> considered 'stale' and gets removed (86400 by default)
>> * |*fileretry*| (numeric) --- number of retries to read a job file
>> (2 by default)
>> * |*filesleep*| (numeric) --- delay in milliseconds between retries
>> above (10 by default)
>> * |*debug*| (0, 1 or 2) --- debug level. There are three of them: 0
>> --- only warnings (default), 1 --- all messages, 2 --- stderr is
>> being redirected to the log file (bad for log parsers, but good
>> for catching problems in Perl jobmanagers)
>>
>> Playing with tick and stateage I managed to cut down waiting time to
>> know
>> from the UI when a job is done or not.
>> But there are some tips on the manual that I couldn't apply:
>>
>> "If your site is running short jobs consider decreasing the tick
>> parameter to 60 and set statefact to a non-zero value (e.g. 3 or 4,
>> check
>> the formula above). In this case stateage may be increased up to 900 and
>> more in order to keep batch system load low."
>>
>> statefact won't let my globus-gma daemon start if it is uncommented on
>> the config file.
>>
>> ¿Is there some other way to reduce the waiting times in the CE for jobs
>> to confirm their state to the UI?
>>
>> Thanks in advance.
>
>
>
> *************************************************************
> * Michel Jouvin Email : [log in to unmask] *
> * LAL / CNRS Tel : +33 1 64468932 *
> * B.P. 34 Fax : +33 1 69079404 *
> * 91898 Orsay Cedex *
> * France *
> *************************************************************
>
>
The globus version is 4.0.3.
Thanks
|