Checking correct error field for retry

[hardillb]

Description

Ensure err.code is checked not err.response.statusCode

It also ensures that the Persistent Storage PVC is only created if it doesn't exist (e.g. don't try and recreate for a suspended instance).

If the PVC is created then it will also add retry logic to the AWS EFS calls if using EFS AccessPoints to store multiple instance storage directories on single EFS (Only really used for FFC)

Related Issue(s)

Checklist

• I have read the contribution guidelines
• Suitable unit/system level tests have been added and they pass
• Documentation has been updated
  • Upgrade instructions
  • Configuration details
  • Concepts
• Changes flowforge.yml?
  • Issue/PR raised on FlowFuse/helm to update ConfigMap Template
  • Issue/PR raised on FlowFuse/CloudProject to update values for Staging/Production
• Link to Changelog Entry PR, or note why one is not needed.

Labels

• Includes a DB migration? -> add the area:migration label

[knolleary]

I'm nervous about the aws-efs logic that now has to query up to 9999 (rather than up to 999) endpoints just to find the one with the most space. Can we use a different heuristic that finds one with 'enough' space so we don't have to poll the whole list every time an instance is created? Or do some caching to remember the top ten with most space and only rescan if necessary?

[Copilot]

Pull request overview

This PR updates the Kubernetes driver and AWS EFS integration to improve retry behavior and avoid unnecessary PVC recreation, aligning error handling with err.code rather than err.response.statusCode.

Changes:

• Add async-retry dependency and use it to retry AWS EFS Describe* calls on throttling.
• Update PVC creation flow to only create a PVC when it does not already exist.
• Adjust Kubernetes API retry logic to check err.code for 429 responses.

Reviewed changes

Copilot reviewed 3 out of 4 changed files in this pull request and generated 3 comments.

File	Description
package.json	Adds async-retry dependency to support new retry behavior.
package-lock.json	Locks async-retry and its transitive dependency retry.
lib/aws-efs.js	Wraps EFS DescribeFileSystems / DescribeAccessPoints in retry logic for throttling cases.
kubernetes.js	Skips PVC creation if it already exists; changes retry condition to use err.code.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[hardillb]

Tested locally and with now wrapping readNamespacedPersistentVolumeClaim we see the error get logged

{"level":"ERROR","time":"2026-03-25T14:27:05.375Z","msg":"[k8s] API call to readNamespacedPersistentVolumeClaim failed. attempt=1/11 statusCode=404  Error: HTTP-Code: 404\nMessage: Unknown API Status Code!\nBody: \"{\\\"kind\\\":\\\"Status\\\",\\\"apiVersion\\\":\\\"v1\\\",\\\"metadata\\\":{},\\\"status\\\":\\\"Failure\\\",\\\"message\\\":\\\"persistentvolumeclaims \\\\\\\"bbee7aba-fd25-471e-bdd1-e73780cc69f8-pvc\\\\\\\" not found\\\",\\\"reason\\\":\\\"NotFound\\\",\\\"details\\\":{\\\"name\\\":\\\"bbee7aba-fd25-471e-bdd1-e73780cc69f8-pvc\\\",\\\"kind\\\":\\\"persistentvolumeclaims\\\"},\\\"code\\\":404}\\n\"\nHeaders: {\"audit-id\":\"c17b9ae3-19a9-4564-9a86-0219ea60816c\",\"cache-control\":\"no-cache, private\",\"connection\":\"close\",\"content-length\":\"284\",\"content-type\":\"application/json\",\"date\":\"Wed, 25 Mar 2026 14:27:05 GMT\",\"x-kubernetes-pf-flowschema-uid\":\"3813c7af-8235-4207-aaf2-88e56242b97a\",\"x-kubernetes-pf-prioritylevel-uid\":\"f12f08f7-9332-46c3-bdc4-ef5fa4ce151b\"}"}

And it correctly created the PVC