[Swift-devel] 0.92.1

Michael Wilde wilde at mcs.anl.gov
Mon Apr 11 16:12:36 CDT 2011


Cool. For Fusion, if you add the line " debug=true" to the end of the file etc/provider-pbs.properties, Swift will save the .submit file for all PBS jobs in ~/.globus/scripts. 


That file gets overwritten when you build from source; in that case you should make the change in your source tree in the source version of that file, under cog/modules/provider-localscheduler/etc. 


- Mike 


----- Original Message -----


I don't think I announced it - but I should have sent a status update on the other tests while I was hammering out the issues with Fusion. Sorry about that. 

David 


On Mon, Apr 11, 2011 at 4:31 PM, Michael Wilde < wilde at mcs.anl.gov > wrote: 




David, thats great to hear about PADS, Argonne servers, and local. Did you announce that, and I missed it? Thats what Ive been waiting to hear. My apologies if I missed an email. 


I think we should release without the Fusion tests. Please swift-devel on any support tickets you file so we can help stay on top of them. 


One possibility is that you're using parameters that will never be met on Fusion. You should post the PBS .submit file from ~/.globus/scripts so we (and Fusion support) can assess why jobs are not running. 


Thanks, 


Mike 









It has been tested on PADS, Argonne systems and locally. I am struggling to get things running on Fusion - I have been getting into a situation where the job will just sit there for days and never execute. I will send an email to their tech support to get some more detail. In the meantime, if anyone wants to take a stab at getting it running on Fusion that would be helpful. There is a test script in tests/providers/local-sge-coasters/002-duplicate-submission-local-sge-coasters.swift. 


Thanks, 

David 


On Mon, Apr 11, 2011 at 3:05 PM, Michael Wilde < wilde at mcs.anl.gov > wrote: 


We were targeting this for last Friday but didnt succeed. And Im not sure we were using the best approach. 

Should we tag a release candidate and test that instead of the raw branch? 

What we said last Wednesday was something like: 

[ ] Sarah: document what systems you test and how you test them 
[ ] Ranger 
[ ] Abe -> will move this to test other TG Systems 
[ ] Beagle 
[ ] David: same for CI/Argonne systems 
[ ] Argonne computer servers 
[ ] PADS 
[ ] Fusion 
[ ] local tests 
[ ] Justin: make 0.92.1 tag 

Sarah and David, where do you stand on the above tests? 

- Mike 










----- Original Message ----- 
> We probably need to make this release sooner rather than later. 
> 
> Mihael 
> 
> _______________________________________________ 
> Swift-devel mailing list 
> Swift-devel at ci.uchicago.edu 
> http://mail.ci.uchicago.edu/mailman/listinfo/swift-devel 

-- 
Michael Wilde 
Computation Institute, University of Chicago 
Mathematics and Computer Science Division 
Argonne National Laboratory 




_______________________________________________ 
Swift-devel mailing list 
Swift-devel at ci.uchicago.edu 
http://mail.ci.uchicago.edu/mailman/listinfo/swift-devel 




-- 

Michael Wilde 
Computation Institute, University of Chicago 
Mathematics and Computer Science Division 
Argonne National Laboratory 





-- 
Michael Wilde 
Computation Institute, University of Chicago 
Mathematics and Computer Science Division 
Argonne National Laboratory 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/swift-devel/attachments/20110411/8dedcb61/attachment.html>


More information about the Swift-devel mailing list