Add ability to pull user defined images in the warm pool

Author: dezmodue

docs/instance_groups.md

@@ -319,18 +319,32 @@ spec:
 You can also specify defaults for all instance groups of type Node or APIServer by setting the `warmPool` field in the cluster spec.
 If warm pools are enabled at the cluster spec level, you can disable them at the instance group level by setting `maxSize: 0`.
 
+### Additional container images
+In some cases it can be convenient to download large container images during the warming phase to reduce the startup time of workloads after a node has been requested and joined the cluster. Additional images can be specified via the `additionalImages` field.
+
+```yaml
+spec:
+  warmPool:
+    additionalImages:
+      - nvcr.io/nvidia/tritonserver:24.10-py3
+      - nvcr.io/nvidia/tritonserver:25.11-vllm-python-py3
+```
+
 ### Lifecycle hook
 
 By default AWS does not guarantee that the kOps configuration will run to completion. Nor that the instance will timely shut down after completion if the instance is allowed to run that long. In order to guarantee this, a lifecycle hook is needed.
 
 **You have to ensure your metadata API is protected if you enable this. If not, any Pod in the cluster will be able to complete the lifecycle hook with the `ABANDONED` result, preventing any instance from ever joining the cluster.**
 
-The following config will enable the lifecycle hook as well as protect the metadata API from abuse:
+By default the lifecycle hook will timeout after 600s. A custom timeout can be set via the `lifecycleHookTimeout` field in the `warmPool` spec. This might be needed if larger additional container images are pulled during the warmup phase so these tasks can finish.
+
+The following config will enable the lifecycle hook, set the timeout to 900s as well as protect the metadata API from abuse:
 
 ```yaml
 spec:
   warmPool:
     enableLifecycleHook: true
+    lifecycleHookTimeout: 900    
   instanceMetadata:
     httpPutResponseHopLimit: 1
     httpTokens: required

k8s/crds/kops.k8s.io_clusters.yaml

@@ -6541,11 +6541,22 @@ spec:
                 description: WarmPool defines the default warm pool settings for instance
                   groups (AWS only).
                 properties:
+                  additionalImages:
+                    description: AdditionalImages is a list of additional container
+                      images to pull into the warm pool instances.
+                    items:
+                      type: string
+                    type: array
                   enableLifecycleHook:
                     description: |-
                       EnableLifecycleHook determines if an ASG lifecycle hook will be added ensuring that nodeup runs to completion.
                       Note that the metadata API must be protected from arbitrary Pods when this is enabled.
                     type: boolean
+                  lifecycleHookTimeout:
+                    description: LifecycleHookTimeout is the timeout for the ASG lifecycle
+                      hook in seconds.
+                    format: int32
+                    type: integer
                   maxSize:
                     description: |-
                       MaxSize is the maximum size of the warm pool. The desired size of the instance group

k8s/crds/kops.k8s.io_instancegroups.yaml

@@ -1189,11 +1189,22 @@ spec:
                 description: WarmPool configures an ASG warm pool for the instance
                   group
                 properties:
+                  additionalImages:
+                    description: AdditionalImages is a list of additional container
+                      images to pull into the warm pool instances.
+                    items:
+                      type: string
+                    type: array
                   enableLifecycleHook:
                     description: |-
                       EnableLifecycleHook determines if an ASG lifecycle hook will be added ensuring that nodeup runs to completion.
                       Note that the metadata API must be protected from arbitrary Pods when this is enabled.
                     type: boolean
+                  lifecycleHookTimeout:
+                    description: LifecycleHookTimeout is the timeout for the ASG lifecycle
+                      hook in seconds.
+                    format: int32
+                    type: integer
                   maxSize:
                     description: |-
                       MaxSize is the maximum size of the warm pool. The desired size of the instance group

pkg/apis/kops/cluster.go

@@ -1106,6 +1106,10 @@ type WarmPoolSpec struct {
 	// EnableLifecyleHook determines if an ASG lifecycle hook will be added ensuring that nodeup runs to completion.
 	// Note that the metadata API must be protected from arbitrary Pods when this is enabled.
 	EnableLifecycleHook bool `json:"enableLifecycleHook,omitempty"`
+	// LifecycleHookTimeout is the timeout for the ASG lifecycle hook in seconds.
+	LifecycleHookTimeout *int32 `json:"lifecycleHookTimeout,omitempty"`
+	// AdditionalImages is a list of additional container images to pull into the warm pool instances.
+	AdditionalImages []string `json:"additionalImages,omitempty"`
 }
 
 func (in *WarmPoolSpec) IsEnabled() bool {

pkg/apis/kops/v1alpha2/cluster.go

@@ -898,4 +898,8 @@ type WarmPoolSpec struct {
 	// EnableLifecycleHook determines if an ASG lifecycle hook will be added ensuring that nodeup runs to completion.
 	// Note that the metadata API must be protected from arbitrary Pods when this is enabled.
 	EnableLifecycleHook bool `json:"enableLifecycleHook,omitempty"`
+	// LifecycleHookTimeout is the timeout for the ASG lifecycle hook in seconds.
+	LifecycleHookTimeout *int32 `json:"lifecycleHookTimeout,omitempty"`
+	// AdditionalImages is a list of additional container images to pull into the warm pool instances.
+	AdditionalImages []string `json:"additionalImages,omitempty"`
 }

pkg/apis/kops/v1alpha2/zz_generated.conversion.go

@@ -862,4 +862,8 @@ type WarmPoolSpec struct {
 	// EnableLifecycleHook determines if an ASG lifecycle hook will be added ensuring that nodeup runs to completion.
 	// Note that the metadata API must be protected from arbitrary Pods when this is enabled.
 	EnableLifecycleHook bool `json:"enableLifecycleHook,omitempty"`
+	// LifecycleHookTimeout is the timeout for the ASG lifecycle hook in seconds.
+	LifecycleHookTimeout *int32 `json:"lifecycleHookTimeout,omitempty"`
+	// AdditionalImages is a list of additional container images to pull into the warm pool instances.
+	AdditionalImages []string `json:"additionalImages,omitempty"`
 }