- 12 Apr, 2017 1 commit
-
-
Morris Jette authored
-
- 11 Apr, 2017 8 commits
-
-
Brian Christiansen authored
If there were no clusters in the system then the dbd would insert a row in the database and then on the next time the dbd would start (with no clusters) it would attempt to insert the row again and fatal. Now it always makes a query against the database and if there are no clusters in the system, then it inserts the current db version into the table.
-
Morris Jette authored
Without this test the macro _entity_update_kv_helper() will use a NULL value for oldvalue and abort Coverity CID 44965
-
Morris Jette authored
Treat the same as having the --local option and only get information from the identified cluster. If sprio --cluster option is set, then implicitly set --local
-
Morris Jette authored
Fix the sprio API so that jobs that are on remote federations only and in multiple partitions do not remove all but one partition's record. Use a hash function to turn the partition name into a number and insure that every unique job ID & partition name are reported by sprio.
-
Morris Jette authored
See Coverity CID 56711
-
Morris Jette authored
If sysconf() fails, it returns -1, which was not being handled properly. Coverity CID 59941
-
Morris Jette authored
-
Morris Jette authored
The sort list function was flawed and there was a bug in the job priority API used by sprio
-
- 10 Apr, 2017 5 commits
-
-
Morris Jette authored
Add sprio --local and --sibling options for use in federation of clusters.
-
Morris Jette authored
-
Morris Jette authored
Move the function to get job priorities from src/sprio to src/api so that we can handle the threading required for federated clusters in a common location in the event this logic is used from other places in the future (e.g. sview). No fundamental change in logic. The logic and definitions were just moved.
-
Morris Jette authored
-
Morris Jette authored
Add sprio -p/--partition option to filter jobs by partition name. Add partition name to job priority factor response message.
-
- 08 Apr, 2017 1 commit
-
-
Morris Jette authored
Modify backfill logic so that bf_max_job_user, bf_max_job_part and bf_max_job_user_part options can all be used independently of each other.
-
- 07 Apr, 2017 1 commit
-
-
Morris Jette authored
Add SchedulerParameters option of bf_max_job_user_part to specifiy the maximum number of jobs per user for any single partition. This differs from bf_max_job_user in that a separate counter is applied to each partition rather than having a single counter per user applied to all partitions.
-
- 06 Apr, 2017 10 commits
-
-
Tim Wickberg authored
-
Tim Wickberg authored
-
Tim Wickberg authored
-
Morris Jette authored
The problem will be solved by using per-partition QOS and setting MaxJobPer[User|Account] on that QOS. bug 3668
-
Danny Auble authored
This was a regression in commit 1967d512.
-
Danny Auble authored
-
Morris Jette authored
Bug 3650
-
Morris Jette authored
This fixes test 7.17 to work with commit a152196f
-
Tim Wickberg authored
And mark thread local variables static to limit scope.
-
Brian Christiansen authored
-
- 05 Apr, 2017 14 commits
-
-
Morris Jette authored
Add SchedulerParameters option of bf_max_job_user_part to specifiy the maximum number of jobs per user for any single partition. This differs from bf_max_job_user in that a separate counter is applied to each partition rather than having a single counter per user applied to all partitions.
-
Danny Auble authored
-
Danny Auble authored
-
Brian Christiansen authored
-
Danny Auble authored
set min cpus based on tasks if we are overcommitting. If we do then it breaks how we overcommit.
-
Danny Auble authored
# Conflicts: # src/plugins/accounting_storage/mysql/as_mysql_convert.c
-
Danny Auble authored
-
Brian Christiansen authored
to remove active sibling from federated job.
-
Brian Christiansen authored
-
Brian Christiansen authored
so that it can be accessible to new function slurm_kill_job_msg.
-
Brian Christiansen authored
to be able to remove specific sibling jobs from a federated job's active siblings.
-
Brian Christiansen authored
-
Brian Christiansen authored
The fed_mgr was trying to only use the locks that it needed so that it could increase parallelism but this could lead to locks. When using the internal locking, you should get all the locks once and then release them.
-
Brian Christiansen authored
-