OpenShift 3.11运行Pod内存限制从2GB扩至4GB的操作及报错解决

Got it, let's break this down step by step. First off, that error you're hitting isn't a bug—it's intentional behavior from OpenShift (and Kubernetes under the hood). Running Pods are mostly immutable; you can't directly edit fields like resource limits on them. Instead, you need to update the Deployment or DeploymentConfig (since this is OpenShift 3.11, both might be in use) that manages the Pod. Here's how to do it via both the web console and oc CLI:

1. Log into your OpenShift 3.11 web console and navigate to the project where your Pod is running.
2. Go to Applications > Deployments (or Applications > DeploymentConfigs if your Pod is managed by a DeploymentConfig) to find the controller that owns your Pod. You can cross-check this by looking at the Pod's "Owner" field in the Pods list.
3. Click on the name of the Deployment/DeploymentConfig to open its details page.
4. Either:
   • Click Actions > Edit Deployment (or "Edit DeploymentConfig") to use the form editor, then find the "Resources" section for your container and update the memory limit from 2Gi to 4Gi.
   • Or switch to the YAML tab, locate the spec.template.spec.containers[].resources.limits.memory field, change its value to 4Gi, then save the changes.
5. After saving, OpenShift will automatically trigger a rolling update: it'll spin up new Pods with the updated memory limit, then terminate the old ones once the new Pods are ready.

First, confirm whether your Pod is managed by a Deployment or DeploymentConfig by running:

oc get pods -o custom-columns=NAME:.metadata.name,OWNER:.metadata.ownerReferences[0].kind,OWNER_NAME:.metadata.ownerReferences[0].name

Option 1: Quick Resource Update (Recommended) #

Use the oc set resources command to directly update the controller's resource limits:

• If it's a Deployment:

  oc set resources deployment/<your-deployment-name> --limits=memory=4Gi
  

• If it's a DeploymentConfig:

  oc set resources dc/<your-dc-name> --limits=memory=4Gi
  

This will immediately trigger a rolling update to deploy new Pods with the 4GB memory limit.

Option 2: Edit the Controller YAML #

If you prefer to edit the YAML directly:

• For a Deployment:

  oc edit deployment/<your-deployment-name>
  

• For a DeploymentConfig:

  oc edit dc/<your-dc-name>
  

In the editor that opens, find the spec.template.spec.containers[].resources.limits.memory line, change 2Gi to 4Gi, save the file, and exit. OpenShift will handle the rest with a rolling update.

Reason: Pod "hjaha" is invalid: spec: Forbidden: pod updates may not change fields other than spec.containers[].image , spec.initContainers[].image , spec.activeDeadlineSeconds or spec.tolerations (only additions to existing tolerations)...

This error exists because Kubernetes (which OpenShift is built on) treats running Pods as immutable for most configuration fields. The only exceptions are a handful of non-critical fields like container images, activeDeadlineSeconds, or adding new tolerations. Resource limits are part of the Pod's core specification, so modifying them directly on a running Pod is blocked. Instead, you have to update the controller that creates and manages the Pod—this ensures that all future Pods (including replacements if the current one fails) use the updated configuration.

内容的提问来源于stack exchange，提问作者william007