Ticket 345

Summary: sview hangs when a large number of jobs are running
Product: Slurm Reporter: Nancy <nancy.kritkausky>
Component: OtherAssignee: Danny Auble <da>
Status: RESOLVED FIXED QA Contact:
Severity: 5 - Enhancement    
Priority: --- CC: da, yiannis.georgiou
Version: 2.6.x   
Hardware: Linux   
OS: Linux   
Site: Atos/Eviden Sites Alineos Sites: ---
Atos/Eviden Sites: --- Confidential Site: ---
Coreweave sites: --- Cray Sites: ---
DS9 clusters: --- HPCnow Sites: ---
HPE Sites: --- IBM Sites: ---
NOAA SIte: --- OCF Sites: ---
Recursion Pharma Sites: --- SFW Sites: ---
SNIC sites: --- Linux Distro: ---
Machine Name: CLE Version:
Version Fixed: Target Release: ---
DevPrio: --- Emory-Cloud Sites: ---
Attachments: sview scalablity patch
scalability for sview

Description Nancy 2013-06-21 07:30:33 MDT 
Created attachment 305 [details]
sview scalablity patch

Scalability problem with SView when there are lot of jobs running and more jobs queued.  We made an improvement related to the memory allocation done when  building the data structures.  We defined a configurable node limit to allocate at a given time.  Then as the window slides, the widget loads a different set of 1000 nodes.  We were only able simulate a large number of nodes.  But it seems to be working quite well.  This work was done by Doug Parisek
Comment 1 Danny Auble 2013-06-21 07:38:32 MDT 
*** Ticket 314 has been marked as a duplicate of this ticket. ***
Comment 2 Danny Auble 2013-06-27 11:47:47 MDT 
Created attachment 316 [details]
scalability for sview

The attached patch appears to increase the speed under heavy load on all tabs.  Let me know if this does not work for you.  This will be in 13.12, but this patch is probably able to work on 2.5 or 2.6.
Comment 3 Danny Auble 2013-06-27 11:48:27 MDT 
If you think things could be improved more than I have already done please reopen.
Comment 4 Nancy 2013-06-28 01:29:04 MDT 
Thanks Danny, We will give it a try.
Nancy