slurmctld
- HPC Charm Team
Channel | Revision | Published | Runs on |
---|---|---|---|
latest/stable | 21 | 14 Jan 2022 | |
latest/candidate | 47 | 17 Oct 2022 | |
latest/edge | 75 | 08 Oct 2024 | |
latest/edge | 59 | 02 Oct 2023 | |
latest/edge | 58 | 24 Aug 2023 |
juju deploy slurmctld
Deploy universal operators easily with Juju, the Universal Operator Lifecycle Manager.
Platform:
-
drain
Drain specified nodes. Example usage: $ juju run-action slurmctld/leader drain nodename=node-[1,2] reason="Updating kernel"
- Params
-
nodename string
The nodes to drain, using the Slurm format, e.g.
node-[1,2]
. -
reason string
Reason to drain the nodes.
- Required
nodename, reason
-
influxdb-info
Get InfluxDB info. This action returns the host, port, username, password, database, and retention policy regarding to InfluxDB.
-
resume
Resume specified nodes. Note: newly added nodes will remain in the
down
state until configured, with thenode-configured
action. Example usage: $ juju run-action slurmctld/leader resume nodename=node-[1,2]- Params
-
nodename string
The nodes to resume, using the Slurm format, e.g.
node-[1,2]
.
- Required
nodename
-
show-current-config
Display the currently used
slurm.conf
. Note: this file only exists inslurmctld
charm and is automatically distributed to all compute nodes by Slurm. Example usage: $ juju run-action slurmctld/leader --format=json --wait | jq .[].results.slurm.conf | xargs -I % -0 python3 -c 'print(%)'