- 27 Oct, 2016 40 commits
-
-
Brian Christiansen authored
It was picking a higher weighted federation over lower weighted federations because it had a earlier starttime. This shouldn't happen because that's what the weights are for. e.g. will_run_resp for fed1: start:2016-10-13T15:19:47 sys_usage:0.00 weight:2 will_run_resp for fed2: start:2016-10-13T15:19:48 sys_usage:0.00 weight:1 will_run_resp for fed3: start:2016-10-13T15:19:48 sys_usage:0.00 weight:1 Earliest cluster:fed1 time:1476393587 now:1476393588 Submitted federated job 67119254 to fed1(self)
-
Brian Christiansen authored
-
Brian Christiansen authored
-
Brian Christiansen authored
fedorigin fedoriginraw fedsiblings fedsiblingsraw
-
Brian Christiansen authored
-
Brian Christiansen authored
-
Brian Christiansen authored
If it exists.
-
Brian Christiansen authored
-
Brian Christiansen authored
-
Brian Christiansen authored
-
Brian Christiansen authored
-
Brian Christiansen authored
-
Brian Christiansen authored
cluster_rec->fed.name will be non-null and empty when the cluster is not part of a federation. Need to check fed.id instead. A fed.id of 0 means the cluster is not part of federation.
-
Brian Christiansen authored
-
Brian Christiansen authored
See previous unreverted commit.
-
Brian Christiansen authored
This reverts commit 2ec92d36a8ad7184897c9a322ba2d9978d2ccdbd.
-
Brian Christiansen authored
This is an example of how to do it. The problem is that select_jobinfo on the job_desc is packed using working_cluster's->plugin_id. job_desc's->select_jobinfo is only used by bluegene and alps code which will eventually go away.
-
Brian Christiansen authored
-
Brian Christiansen authored
sbatch will choose the federation or individual cluster with the fast startime. A willrun to a federation will return the fastest start time of all clusters in a federation.
-
Brian Christiansen authored
with the sbatch -M<clusters> option. A fed will run will the fastest time of all siblings in a federation.
-
Brian Christiansen authored
-
Brian Christiansen authored
Was moved in 5cfc577a
-
Brian Christiansen authored
-
Brian Christiansen authored
More to come. This sets up the controller side.
-
Brian Christiansen authored
-
Brian Christiansen authored
-
Brian Christiansen authored
-
Brian Christiansen authored
It could have state saved a null federation.
-
Brian Christiansen authored
instead of one at a time. Saves ~20 seconds.
-
Brian Christiansen authored
If all of the clusters get updated at the same time, the could get in a state where they are waiting on each other to respond and will eventually timeout and will then will reconnect. By spacing out the clusters being added this helps prevent them from talking to everyone at the same time.
-
Brian Christiansen authored
-
Brian Christiansen authored
-
Brian Christiansen authored
-
Brian Christiansen authored
-
Brian Christiansen authored
-
Brian Christiansen authored
-
Brian Christiansen authored
get_next_job_id() didn't take into consideration job_ids that already be taken by other jobs like set_job_id() did.
-
Brian Christiansen authored
Send the existing packed buffer that has the job_desc to the sibling. The sibling will unpack it on the other side. This prevents having to pack the job_desc for each willrun/allocation to each sibling.
-
Brian Christiansen authored
squeue --fedtrack
-
Brian Christiansen authored
squeue long options: fedorigin, fedoriginraw, fedsiblings and fedsiblingsraw.
-