[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[HTCondor-users] HTCondor 8.3.1 memory handling problem




Hello,

I run HTCondor condor-8.3.1 on Fedora 20.  It works as it should, but I get
the following warning/error for each job: (StarterLog.slot1_1)

----------
[...]
11/06/14 10:48:17 (pid:2352) Create_Process succeeded, pid=2353
11/06/14 10:48:17 (pid:2352) Limitting memory usage to 7918845952 bytes
11/06/14 10:48:17 (pid:2352) Limitting memsw usage to 25475178496 bytes
11/06/14 10:48:17 (pid:2352) Unable to commit memsw limit for /condor/condor_usr_data_condor_execute_slot1_1@xxxxxxxxxxxxxxxxxxxxxx : 50016 Invalid argument 11/06/14 10:48:17 (pid:2352) Unable to commit CPU shares for /condor/condor_usr_data_condor_execute_slot1_1@xxxxxxxxxxxxxxxxxxxxxx: 50016 Invalid argument
11/06/14 10:49:31 (pid:2352) Process exited, pid=2353, status=0
[...]
----------

Apart from the spelling error in 'limitting', this does not sound nice ...

This was not there in 8.3.0, so I suspect it has to do with the new feature
mentioned in the release notes for 8.3.1:
(see http://research.cs.wisc.edu/htcondor/manual/v8.3.1/10_3Development_Release.html)

----------
If cgroups are enabled on Linux platforms, the amount of swap space used by a job is now limited to the size specified by the machine ClassAd attribute VirtualMemory for the slot that the job is running on. (Ticket #4417).
----------

Is anyone else seeing this, or is it an artefact of my running the 8.3.1 RHEL7
packages on Fedora? (The official Fedora package is still at 8.1.1 from October last year).

Greetings, Bert.

--
"Das einzige Gegenargument gegen eine dumme Stirn ist eine geballte Faust."
F. Nietzsche.