Re: [HTCondor-users] Misreporting Nodes

[Michael Pelletier <michael.v.pelletier@xxxxxxxxxxxx>]

Hi Justin,

 

This is how âpartitionableâ slots work. When a job comes in, a slot of the requested size is pulled out of the machineâs main slot and set up as a âdynamicâ slot, which lasts for the duration of the job (or the job cluster for the lease duration).

 

This is a more flexible approach than the previous âstaticâ slots configuration where machines were divided up into fixed-size slots.

 

Back when HTCondor was first developed, machines only had one CPU core, and the partitionable/dynamic approach handles allocating resources of machines with multiple CPU cores by treating each group of resources for a given job as a âmachine.â

 

Michael V. Pelletier
Information Technology
Digital Transformation & Innovation
Integrated Defense Systems
Raytheon Company

 

From: HTCondor-users <htcondor-users-bounces@xxxxxxxxxxx> On Behalf Of Justin Fisher
Sent: Wednesday, March 11, 2020 11:51 AM
To: HTCondor-Users Mail List <htcondor-users@xxxxxxxxxxx>
Subject: [External] [HTCondor-users] Misreporting Nodes

 

Hi.

 

When idle, a condor_status reports 48 nodes on my system, which is correct. When in use though it reports 96 nodes of which 48 are in use. Obviously the nodes are being fully utilized, so this is really no big deal - but I must have some configuration error somewhere. Any ideas where I should look?

 

slot24@xxxxxxxxxxxxxxxxxxx   LINUX      X86_64 Unclaimed Idle      0.000 1699  0+00:00:33
slot24_1@xxxxxxxxxxxxxxxxxxx LINUX      X86_64 Claimed   Busy      0.690 4096  0+00:00:04

--
Kind regards,

 

Justin Fisher.