Ticket 450

Summary: Select Plugin Omits an important option to the nodehealth checker binary
Product: Slurm Reporter: Jason Sollom <jasons>
Component: Cray ALPSAssignee: Danny Auble <da>
Status: RESOLVED FIXED QA Contact:
Severity: 3 - Medium Impact    
Priority: --- CC: da
Version: 14.03.x   
Hardware: Linux   
OS: Linux   
Site: CRAY Slinky Site: ---
Alineos Sites: --- Atos/Eviden Sites: ---
Confidential Site: --- Coreweave sites: ---
Cray Sites: --- DS9 clusters: ---
Google sites: --- HPCnow Sites: ---
HPE Sites: --- IBM Sites: ---
NOAA SIte: --- NoveTech Sites: ---
Nvidia HWinf-CS Sites: --- OCF Sites: ---
Recursion Pharma Sites: --- SFW Sites: ---
SNIC sites: --- Tzag Elita Sites: ---
Linux Distro: --- Machine Name:
CLE Version: Version Fixed:
Target Release: --- DevPrio: ---
Emory-Cloud Sites: ---

Description Jason Sollom 2013-10-09 09:17:18 MDT 
When the select_cray plugin calls xtcleanup_after, it omits the '-m' option.  This option tells NHC whether it is running at the end of the job step or the end of the job.  The -m option takes a non-option argument string either "application" or "reservation".

For the end of the job step, the option should be

-m application

For the end of the job, the option should be

-m reservation

Please upgrade the select_cray plugin to use the -m option.

NHC will not be properly invoked for the end of the job until this change is made.  

This will inhibit the current testing that is going on.
Comment 1 Danny Auble 2013-10-09 09:26:13 MDT 
Thanks for dropping knowledge on us Jason ;).

I apparently though (foolishly) the -a or -r options would of been sufficient.

This is fixed in 2aa193ef2c366adf39fafeb9c1284d5367cf2118
Comment 2 Danny Auble 2013-10-09 09:54:56 MDT 
Do not use the previous plugin, the xfrees were not updated to reflect the change in position of allocated memory.

I am waiting on a response to bug 451 to commit a fix since everything changes again.