What is the nature and description of the request?

Currently when a Controller task pod is removed/deleted/changed/etc but the Controller web pods are not restarted along with that, the old task pod is still reported in the metrics endpoint since the information in redis is not cleared for the old task pod.

Why does the customer need this? (List the business requirements here)

So that incorrect metrics aren't actioned upon when ingesting the metrics into other tools.

How would you like to achieve this? (List the functional requirements here)

There's a couple ideas:

• A periodic check to confirm that the information held for metrics matches what's currently out there
• Or a step on deprovisioning the old task pod that also removes information from redis for that pod.