- 12 Oct, 2016 10 commits
-
-
Brian Christiansen authored
The for loop wasn't resetting the job_start_time to the reservation's end_time. This was taken out in 06dc5fa7.
-
Brian Christiansen authored
-
Brian Christiansen authored
Changed in df70b651
-
Brian Gilmer authored
-
Morris Jette authored
-
Morris Jette authored
Preserve non-KNL node features when updating the KNL node features for a multi-node job in which the non-KNL node features vary by node.
-
Morris Jette authored
node_features/knl_cray plugin: If the reconfiguration of nodes for an interactive job fails, kill the job (it can't be requeued like a batch job).
-
Morris Jette authored
-
Morris Jette authored
Execute "capmc node_status" at more frequent intervals to handle nodes getting added or removed from the system using Cray tools (i.e. try to keep Slurm and Cray software better synchronized).
-
Morris Jette authored
node_features/knl_cray plugin: Add separate thread to interact with capmc in response to unexpected node reboots. bug 3153
-
- 11 Oct, 2016 11 commits
-
-
Alejandro Sanchez authored
bug 3091
-
Morris Jette authored
-
Morris Jette authored
-
Morris Jette authored
-
Morris Jette authored
bug 3155
-
Morris Jette authored
Prevent possible divide by zero in select/cons_res if a node's board count is higher than it's socket count. bug 3155
-
Morris Jette authored
If a node's socket or core count are changed at registration time (e.g. a KNL node's NUMA mode is changed), change it's board count to match. bug 3155
-
Morris Jette authored
Cray: The slurmd can manipulate the socket/core/thread values reported based upon the configuration. The logic failed to consider select/cray with SelectTypeParameters=other_cons_res as equivalent to select/cons_res. bug 3155
-
Tim Wickberg authored
-
Tim Wickberg authored
abs() should not be used on long long variables as it would truncate if strictly confirming to C99. Use llabs() instead. Fix to commit 2aefc66b.
-
Tim Wickberg authored
-
- 10 Oct, 2016 1 commit
-
-
Tim Wickberg authored
Inadvertently broken by commit 3251406c.
-
- 09 Oct, 2016 1 commit
-
-
Morris Jette authored
These tests fail with a configuration of CR_ONE_TASK_PER_CORE. The tests are being disabled now for such a configuration, but could be modified to run in such an environment in the future.
-
- 07 Oct, 2016 9 commits
-
-
Tim Wickberg authored
No functional change.
-
Morris Jette authored
Expand test to run on various allocation sizes
-
Morris Jette authored
Replaced by test1.114
-
Morris Jette authored
-
Morris Jette authored
Previous logic would result in sub-optimal job allocations and under some conditions invalid memory references resulting in the slurmctld daemon crashing or aborting. bug 2732
-
Morris Jette authored
-
Morris Jette authored
-
Morris Jette authored
-
Morris Jette authored
Correct SchedulerParameters=bf_busy_nodes logic with respect to the job's minimum node count. Previous logic would not decremement counter in two locations and reject valid job request for not reaching minimum node count.
-
- 06 Oct, 2016 8 commits
-
-
Danny Auble authored
always just AccountingPolicy.
-
Tim Wickberg authored
-
Morris Jette authored
-
Danny Auble authored
daf91a80
-
Danny Auble authored
-
Danny Auble authored
configure.
-
Morris Jette authored
Add description of how to test for disabled Cray nodes and avoid trying to use them in job allocations. Access to Cray required to flesh out the code and test.
-
Morris Jette authored
-