Persist Connector Information even after Kubernetes pods restart

[ArunavNobroker]

I am currently working on deploying a strimzi kafka connect cluster in GKE. As i was testing it I noticed that everytime the pod restarts the connector in that pod is lost. As i plan to use spot nodes for the connect cluster can someone please help me in how I can configure the cluster such that when a pod is restarted the connectors are restarted as well

[scholzj]

What does it mean that the Connector is lost? The connectors are definitely bormally "persisted" accross rolling updates. But depending on the Strimzi version, they might not end up rescheduled on the same pod. So if you somehow loose the connecotr, you should share more info what does it mean, logs and configs.

…

On Sat, Nov 11, 2023, 09:19 ArunavNobroker ***@***.***> wrote: I am currently working on deploying a strimzi kafka connect cluster in GKE. As i was testing it I noticed that everytime the pod restarts the connector in that pod is lost. As i plan to use spot nodes for the connect cluster can someone please help me in how I can configure the cluster such that when a pod is restarted the connectors are restarted as well — Reply to this email directly, view it on GitHub <#9343>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ABLFOR5YQN5J7OR5ZJRFSNDYD4YITAVCNFSM6AAAAAA7HDALQKVHI2DSMVQWIX3LMV43ERDJONRXK43TNFXW4OZVHAZTQMZWGI> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>

[ArunavNobroker]

If I delete the pod or the node where the pod is running gets restarted which will happen in spot vm instances, the connector is lost.

[ArunavNobroker]

Nevermind when I am creating a connector though kubectl it's being persisted even when pod restarts. Only if I create it though the pod ip it's getting lost