Skip to main content

Manage a Slurm Cluster

Everything after creation happens on the cluster details page: ten tabs and an Actions menu. This is the map.


Open a Cluster

  1. Log in to the TIR console and select your project.
  2. Click Training Cluster in the sidebar. The list page is headed Manage Training Clusters.
  3. Click the cluster name.

The list shows Cluster Name, Created By, Plan, No of Nodes, Created At, a Connect shortcut, Status, and an Actions (⋮) menu per row. It refreshes itself every few seconds while any cluster is Creating or Deleting.


The Page Header

ControlWhat it does
Select Training ClusterSwitch to another cluster without going back to the list
Status pillGreen for Running, red for Failed / Terminated, amber for transitional states
Refresh iconReload the current tab. The details page does not auto-refresh
ConnectOpens the SSH panel with copyable commands
Restart WorkersShortcut for Restart All Workers. Disabled unless Running
Actions (⋮)Every cluster action

The Tabs

TabWhat it is forPage
DetailsName, image version, status, node count, plan, and how to connectbelow
Cluster OverviewGPU and job summary, node health, and the Slurm partition tableCluster Overview
NodesPer-node and per-GPU DCGM metrics, health filters, XID errors, node rebootNodes
JobsLive squeue with state counters and filtersJobs
MonitoringGPU and scheduler charts over selectable time rangesMonitoring
AlertsAttach alerts to the clusterMonitoring
LogsController, worker and login-node logsLogs
VolumesMount and unmount PFS, SFS and datasetsStorage
Network & SecurityPublic IP, VPC IP, and security groupsNetwork and security
Login User ManagementPer-user SSH identities and groupsLogin User Management
Most tabs need a Running cluster

On a cluster that is not Running, tabs such as Cluster Overview, Nodes and Login User Management show a message like "Cluster overview is available when the cluster is running." The Details tab always works, which is where you check status and plan on a Failed or Terminated cluster. Tab availability also varies slightly by region.


Details Tab

Three collapsible sections.

Cluster Details

FieldMeaning
Cluster NameThe name you gave it
Image VersionThe cluster's Slurm image, e.g. v2-slurm25.11-ubuntu24.04. Change it with Update Image
StatusCurrent state
Number of NodesWorker nodes. Change it with Scale Cluster
Created ByWho created it
Created AtWhen

Plan Details

FieldMeaning
Plan NameThe plan or Private Cluster
CPU / MemoryPer-node resources
Plan TypeOn-demand, committed, or private cluster
Private Cluster NameThe reserved cluster it draws from, where applicable

Connection Details

FieldMeaning
SSH KeysThe keys with root access, as chips. Change with Update SSH Keys
SSH Commandssh root@<ip>, with a copy button
The image version tells you what Slurm you have

v2-slurm25.11-ubuntu24.04 means Slurm 25.11 on Ubuntu 24.04. Worth knowing when you read the SchedMD documentation — directives and flags do change between Slurm releases.


Common Tasks

I want to…Go to
See why my job is not runningJobs tab
Check whether the GPUs are actually busyMonitoring
Find a broken GPUNodes
Add more nodesScale Cluster
Add a partitionManage partitions
Change a scheduler settingExtra slurm.conf
Run setup before every jobProlog and epilog scripts
Attach more storageStorage and volumes
Give a colleague accessLogin User Management
Pin the cluster's IPNetwork and security
Stop paying for itTerminate Cluster
Work out why a control is missingAction availability

Last updated on September 10, 2026.