创建OpenShift模板应用（NodeJS）：构建一直Pending状态求助

Hey there, let’s work through this issue together—since you’ve already followed the guide but still hit a wall with empty imagestreams and stuck builds, let’s break down some common missing checks:

• Validate Imagestream Health:
  First, confirm your imagestream exists with oc get imagestreams. If it’s there, run oc describe imagestream <your-imagestream-name> and scan the Events section. Silent failures (like network blocks preventing image pulls, invalid image references, or permission gaps) often show up here even if you followed the step-by-step guide.

• Double-Check Build Config Triggers:
  Make sure your build config is properly linked to the imagestream. Run oc describe buildconfig <your-buildconfig-name> and look at the Triggers section. If you’re using an imagestream trigger, verify the from field points to the exact imagestream name and tag—typos here are super common and easy to miss.

• Test Manual Image Import:
  Skip the automated flow for a minute and manually import an image to rule out import pipeline issues. Use this command (replace placeholders with your details):

  oc import-image <your-imagestream-name>:<target-tag> --from=<external-registry>/<image-name>:<tag> --confirm
  

  If this fails, it’ll spit out a clear error—maybe you’re missing a pull secret for a private registry, the cluster can’t reach the external registry due to firewalls, or the image tag doesn’t exist.

• Verify Builder Service Account Permissions:
  The default builder service account needs permissions to pull images from imagestreams and push built images. Check its roles with oc describe sa builder. If it’s missing image-puller, image-pusher, or edit roles, add them with:

  oc policy add-role-to-user edit system:serviceaccount:<your-namespace>:builder
  

• Inspect Build Pod Status:
  Even if the build shows "pending", sometimes the underlying pod has clues. Run oc get pods -l build=<your-build-name>—if a pod exists, describe it with oc describe pod <pod-name> to spot issues like insufficient cluster resources, scheduling conflicts, or missing secrets. If no pod exists, your build config might not be configured to spawn one, or the OpenShift build controller has a glitch.

• Check Internal Registry Health:
  If you’re using OpenShift’s internal image registry, confirm it’s running. Run oc get pods -n openshift-image-registry—if registry pods aren’t in a Ready state, that’ll block all image imports and build workflows.

I’ve run into cases where tiny oversights (like a misspelled image tag or a forgotten pull secret for a private repo) caused this exact problem, even after following official guides. Starting with the manual import test is usually the fastest way to uncover hidden errors.

内容的提问来源于stack exchange，提问作者Gabriel Muniz Antonio