A node being lent says when a replica with nowhere to go is stopped
CI / test (push) Successful in 9m37s
CI / test (push) Successful in 9m37s
Tried on the PoC: lending cuda01, the one GPU node, its model had nowhere else to go and waited out the drain's five-minute grace, while ctl slurm status said only "1 replica(s) still to move off". It now says the time that replica is stopped at. Also recorded: Slurm on a GPU node, tried on cuda01 (gres, ConstrainDevices, accounting of gres/gpu). Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
This commit is contained in:
1 parent
4b4306bb6a
commit
c3a75dcf3f
4 files changed
+56
-5
No files matched your search
@@ -12336,6 +12336,11 @@
|
||||
"replicas": {
|
||||
"type": "integer"
|
||||
},
|
||||
"stops_by": {
|
||||
"type": "string",
|
||||
"format": "date-time",
|
||||
"description": "StopsBy is, while it is being lent, when a replica still on it is stopped even with nowhere else to go: the drain's grace after the lend began. Zero when not known (a leader that took over meanwhile)."
|
||||
},
|
||||
"slurm": {
|
||||
"$ref": "#/components/schemas/slurmsvc.Reported"
|
||||
},
|
||||
|
||||
Reference in new issue
Block a user