feat: allow to specify nodespec

vsoch · vsoch · commit b797793b39a4 · 2026-04-22T21:01:00.000-07:00
Signed-off-by: vsoch &lt;vsoch@users.noreply.github.com&gt;
diff --git a/api/v1alpha1/slurm_types.go b/api/v1alpha1/slurm_types.go
@@ -168,6 +168,11 @@ type Node struct {
 	// +optional
 	WorkingDir string `json:"workingDir,omitempty"`
 
+	// Node specification. Leave empty for testing cluster
+	// This does not include hostlist (generated automatically)
+	// +optional
+	Nodespec string `json:"nodespec,omitempty"`
+
 	// PullAlways will always pull the container
 	// +optional
 	PullAlways bool `json:"pullAlways"`
@@ -238,6 +243,10 @@ func (s *Slurm) Validate() bool {
 		s.Spec.ClusterName = "linux"
 	}
 
+	// Default node spec
+	if s.Spec.Node.Nodespec == "" {
+		s.Spec.Node.Nodespec = "RealMemory=1000 CPUs=1 State=UNKNOWN"
+	}
 	// Along with a username and password
 	if s.Spec.Database.DatabaseName == "" {
 		s.Spec.Database.DatabaseName = "slurm_acct_db"
diff --git a/config/crd/bases/flux-framework.org_slurms.yaml b/config/crd/bases/flux-framework.org_slurms.yaml
@@ -65,6 +65,11 @@ spec:
                     default: ghcr.io/converged-computing/slurm
                     description: Image to use for slurm
                     type: string
+                  nodespec:
+                    description: |-
+                      Node specification. Leave empty for testing cluster
+                      This does not include hostlist (generated automatically)
+                    type: string
                   ports:
                     description: |-
                       Ports to be exposed to other containers in the cluster
@@ -188,6 +193,11 @@ spec:
                     default: ghcr.io/converged-computing/slurm
                     description: Image to use for slurm
                     type: string
+                  nodespec:
+                    description: |-
+                      Node specification. Leave empty for testing cluster
+                      This does not include hostlist (generated automatically)
+                    type: string
                   ports:
                     description: |-
                       Ports to be exposed to other containers in the cluster
@@ -266,6 +276,11 @@ spec:
                     default: ghcr.io/converged-computing/slurm
                     description: Image to use for slurm
                     type: string
+                  nodespec:
+                    description: |-
+                      Node specification. Leave empty for testing cluster
+                      This does not include hostlist (generated automatically)
+                    type: string
                   ports:
                     description: |-
                       Ports to be exposed to other containers in the cluster
diff --git a/controllers/slurm/templates.go b/controllers/slurm/templates.go
@@ -50,6 +50,7 @@ type ConfigTemplate struct {
 	ControlHost  string
 	DatabaseHost string
 	Hostlist     string
+	Nodespec     string
 }
 
 // combineTemplates into one "start"
@@ -124,6 +125,7 @@ func generateConfig(cluster *api.Slurm, startTemplate string) (string, error) {
 		DatabaseHost: database,
 		DaemonHost:   daemon,
 		Hostlist:     generateHostlist(cluster),
+		Nodespec:     cluster.Spec.Worker.Nodespec,
 	}
 
 	// Wrap the named template to identify it later
diff --git a/controllers/slurm/templates/slurm.conf b/controllers/slurm/templates/slurm.conf
@@ -89,8 +89,7 @@ AccountingStoragePort=6819
 #AccountingStorageUser=
 #
 # COMPUTE NODES
-# TODO customize memory here?
-NodeName={{ .Hostlist }} RealMemory=1000 State=UNKNOWN
+NodeName={{ .Hostlist }} {{ .Spec.Node.Nodespec }}
 #
 # PARTITIONS
 PartitionName=normal Default=yes Nodes={{ .Hostlist }} Priority=50 DefMemPerCPU=500 Shared=NO MaxNodes=2 MaxTime=5-00:00:00 DefaultTime=5-00:00:00 State=UP
diff --git a/examples/dist/slurm-operator-arm.yaml b/examples/dist/slurm-operator-arm.yaml
@@ -77,6 +77,11 @@ spec:
                     default: ghcr.io/converged-computing/slurm
                     description: Image to use for slurm
                     type: string
+                  nodespec:
+                    description: |-
+                      Node specification. Leave empty for testing cluster
+                      This does not include hostlist (generated automatically)
+                    type: string
                   ports:
                     description: |-
                       Ports to be exposed to other containers in the cluster
@@ -200,6 +205,11 @@ spec:
                     default: ghcr.io/converged-computing/slurm
                     description: Image to use for slurm
                     type: string
+                  nodespec:
+                    description: |-
+                      Node specification. Leave empty for testing cluster
+                      This does not include hostlist (generated automatically)
+                    type: string
                   ports:
                     description: |-
                       Ports to be exposed to other containers in the cluster
@@ -277,6 +287,11 @@ spec:
                     default: ghcr.io/converged-computing/slurm
                     description: Image to use for slurm
                     type: string
+                  nodespec:
+                    description: |-
+                      Node specification. Leave empty for testing cluster
+                      This does not include hostlist (generated automatically)
+                    type: string
                   ports:
                     description: |-
                       Ports to be exposed to other containers in the cluster
diff --git a/examples/dist/slurm-operator.yaml b/examples/dist/slurm-operator.yaml
@@ -77,6 +77,11 @@ spec:
                     default: ghcr.io/converged-computing/slurm
                     description: Image to use for slurm
                     type: string
+                  nodespec:
+                    description: |-
+                      Node specification. Leave empty for testing cluster
+                      This does not include hostlist (generated automatically)
+                    type: string
                   ports:
                     description: |-
                       Ports to be exposed to other containers in the cluster
@@ -200,6 +205,11 @@ spec:
                     default: ghcr.io/converged-computing/slurm
                     description: Image to use for slurm
                     type: string
+                  nodespec:
+                    description: |-
+                      Node specification. Leave empty for testing cluster
+                      This does not include hostlist (generated automatically)
+                    type: string
                   ports:
                     description: |-
                       Ports to be exposed to other containers in the cluster
@@ -277,6 +287,11 @@ spec:
                     default: ghcr.io/converged-computing/slurm
                     description: Image to use for slurm
                     type: string
+                  nodespec:
+                    description: |-
+                      Node specification. Leave empty for testing cluster
+                      This does not include hostlist (generated automatically)
+                    type: string
                   ports:
                     description: |-
                       Ports to be exposed to other containers in the cluster

Original file line number	Diff line number	Diff line change
`@@ -50,6 +50,7 @@ type ConfigTemplate struct {`
`50`	`50`	`ControlHost string`
`51`	`51`	`DatabaseHost string`
`52`	`52`	`Hostlist string`
	`53`	`+ Nodespec string`
`53`	`54`	`}`
`54`	`55`
`55`	`56`	`// combineTemplates into one "start"`
`@@ -124,6 +125,7 @@ func generateConfig(cluster *api.Slurm, startTemplate string) (string, error) {`
`124`	`125`	`DatabaseHost: database,`
`125`	`126`	`DaemonHost: daemon,`
`126`	`127`	`Hostlist: generateHostlist(cluster),`
	`128`	`+ Nodespec: cluster.Spec.Worker.Nodespec,`
`127`	`129`	`}`
`128`	`130`
`129`	`131`	`// Wrap the named template to identify it later`
Original file line number	Diff line number	Diff line change
`@@ -89,8 +89,7 @@ AccountingStoragePort=6819`
`89`	`89`	`#AccountingStorageUser=`
`90`	`90`	`#`
`91`	`91`	`# COMPUTE NODES`
`92`		`-# TODO customize memory here?`
`93`		`-NodeName={{ .Hostlist }} RealMemory=1000 State=UNKNOWN`
	`92`	`+NodeName={{ .Hostlist }} {{ .Spec.Node.Nodespec }}`
`94`	`93`	`#`
`95`	`94`	`# PARTITIONS`
`96`	`95`	`PartitionName=normal Default=yes Nodes={{ .Hostlist }} Priority=50 DefMemPerCPU=500 Shared=NO MaxNodes=2 MaxTime=5-00:00:00 DefaultTime=5-00:00:00 State=UP`