Skip to content

Update sky map job condor memory request

Cody Messick requested to merge cody.messick/gwcelery:skymap_memory_request into main

Updating the memory request in the hope that it will solve playground skymap jobs being held. We've been seeing hold messages like this for a while

82212.0   emfollow-playg  2/24 22:11 Error from slot2@node765.cluster.ldas.cit: Job has gone over memory limit of 8192 megabytes. Peak usage: 7820 megabytes.
82212.1   emfollow-playg  2/24 20:23 Error from slot2@node746.cluster.ldas.cit: Job has gone over memory limit of 8192 megabytes. Peak usage: 7829 megabytes.
82212.2   emfollow-playg  2/24 21:04 Error from slot2@node2066.cluster.ldas.cit: Job has gone over memory limit of 8192 megabytes. Peak usage: 7823 megabytes.
82212.7   emfollow-playg  2/24 22:10 Error from slot2@node2081.cluster.ldas.cit: Job has gone over memory limit of 8192 megabytes. Peak usage: 7814 megabytes.
82212.9   emfollow-playg  2/25 23:11 Error from slot2@node2073.cluster.ldas.cit: Job has gone over memory limit of 8192 megabytes. Peak usage: 7819 megabytes.
82212.11  emfollow-playg  2/25 21:59 Error from slot2@node2087.cluster.ldas.cit: Job has gone over memory limit of 8192 megabytes. Peak usage: 7816 megabytes.
82212.13  emfollow-playg  2/24 17:42 Error from slot2@node2075.cluster.ldas.cit: Job has gone over memory limit of 8192 megabytes. Peak usage: 7832 megabytes.

Talking with Leo, each of these jobs run on their own machine that has 48 cores and 256 GB of ram, so 96 GB of ram shouldn't be a problem.

Edit: We decided to try 16 first

Edited by Cody Messick

Merge request reports