perfsonar-dev - Re: [GN2-JRA1] [Ticket#2007103110000016] Problem with SWITCH RRD MA
Subject: perfsonar development work
List archive
- From: Jason Zurawski <>
- To: Nina Jeliazkova <>
- Cc: Roman Lapacz <>, "" <>
- Subject: Re: [GN2-JRA1] [Ticket#2007103110000016] Problem with SWITCH RRD MA
- Date: Mon, 03 Dec 2007 11:52:25 -0500
- Openpgp: id=B94D59A6; url=http://people.internet2.edu/~zurawski/key.txt
- Organization: Internet2
Nina;
Bumping to this to pS-Dev, since this is something interesting to mention...
>>> strangely enough, after restarting PerfsonarUI and changing nothing
>>> else, it seems to be working.
>>> It is even reasonably fast, UNLESS:
>>>
>>> if you at any time select "retrieve all" it will remain slow for a
>>> loooooooong time after that, even if you issue a stop.
>>>
>>>
>> I installed a service with your metadata config file. My local machine
>> is 2.0 GHz with 512 MB memory and getting all metadatas took about 50
>> secs (I was using simple java command-line app).
>>
> Perfsonar UI "Retrieve all" sends not only a single metadata request,
> but after it a number of setupdata requests in order to retrieve some
> statistics. In case of SWITCH MA with 343 interfaces, this results in
> sending 43 SetupData requests simultaneously (according to the default
> settings). These set of requests might be the reason of server being
> fully loaded.
>
I think the "retrieve all" is a good thing to have, but particularly
when testing large deployments (ESnet has 1000s of interfaces for
example) it would be good to have functionality to do the metadata
request to populate pS-UI, then selectively retrieve interface info for
data requests instead of automatically getting everythihng. Working on
a Mac mini at SC, it was impossible to run pS-UI due to the volume of
data generated by certain services (also caused repeated Heap Overflows).
> An usefull performance test will be to send several tens of requests
> simultaneously from a simple command line (or SOAPUI) client .
>
> Stop doesn't stop connections (there is no way in Axis1 to interrupt
> connection), it just tells the client side not to wait for them.
>
Just a suggestion, stop is probably a poor word to use (I realized how
it worked before, but newer users may not understand the distinction).
I would expect 'abort' to do the behavior you have described.
-jason
- Re: [GN2-JRA1] [Ticket#2007103110000016] Problem with SWITCH RRD MA, Jason Zurawski, 12/03/2007
- Re: [GN2-JRA1] [Ticket#2007103110000016] Problem with SWITCH RRD MA, Nina Jeliazkova, 12/04/2007
- Re: [pS-dev] Re: [GN2-JRA1] [Ticket#2007103110000016] Problem with SWITCH RRD MA, Jason Zurawski, 12/04/2007
- Re: [pS-dev] Re: [GN2-JRA1] [Ticket#2007103110000016] Problem with SWITCH RRD MA, Roman Lapacz, 12/04/2007
- Re: [pS-dev] Re: [GN2-JRA1] [Ticket#2007103110000016] Problem with SWITCH RRD MA, Chris Welti, 12/04/2007
- Re: [pS-dev] Re: [GN2-JRA1] [Ticket#2007103110000016] Problem with SWITCH RRD MA, Nina Jeliazkova, 12/04/2007
- Re: [pS-dev] Re: [GN2-JRA1] [Ticket#2007103110000016] Problem with SWITCH RRD MA, Chris Welti, 12/04/2007
- Re: [pS-dev] Re: [GN2-JRA1] [Ticket#2007103110000016] Problem with SWITCH RRD MA, Loukik Kudarimoti, 12/04/2007
- Re: [pS-dev] Re: [GN2-JRA1] [Ticket#2007103110000016] Problem with SWITCH RRD MA, Nina Jeliazkova, 12/04/2007
- Re: [pS-dev] Re: [GN2-JRA1] [Ticket#2007103110000016] Problem with SWITCH RRD MA, Chris Welti, 12/04/2007
- Re: [pS-dev] Re: [GN2-JRA1] [Ticket#2007103110000016] Problem with SWITCH RRD MA, Jason Zurawski, 12/04/2007
- Re: [pS-dev] Re: [GN2-JRA1] [Ticket#2007103110000016] Problem with SWITCH RRD MA, Nicolas Simar, 12/04/2007
- Re: [pS-dev] Re: [GN2-JRA1] [Ticket#2007103110000016] Problem with SWITCH RRD MA, Nina Jeliazkova, 12/04/2007
- Re: [pS-dev] Re: [GN2-JRA1] [Ticket#2007103110000016] Problem with SWITCH RRD MA, Jason Zurawski, 12/04/2007
- Re: [GN2-JRA1] [Ticket#2007103110000016] Problem with SWITCH RRD MA, Nina Jeliazkova, 12/04/2007
Archive powered by MHonArc 2.6.16.