LD relay fails to reconnect to Redis after Redis restarts

[carlosjgp]

LD relay fails to reconnect to Redis after Redis restarts

[cwaldren-ld]

Hi @carlosjgp, really appreciate this report. We'll look into it and get back to you.

Filed internally as 178424.

[keelerm84]

@carlosjgp in your report when you mention "after Redis has restarted", how exactly were you simulating that? Scaling the stateful set back down and up again shows the relay reconnecting appropriately.

[carlosjgp]

@carlosjgp in your report when you mention "after Redis has restarted", how exactly were you simulating that? Scaling the stateful set back down and up again shows the relay reconnecting appropriately.

@keelerm84 you can just terminate the POD

kubectl delete po -n <namespace> <redis master pod>

or as you mention

kubectl -n <namespace> scale statefulset <redis-master-sts> --replicas 0 

and after a while --replicas 1

[keelerm84]

Hmmm, I have tried reproducing this with both options, and each worked as expected with the relay re-establishing connection to the redis instance.

You state that the relay fails to reconnect; looking at yours logs, can you confirm if it successfully establishes connection initially before restarting redis?

How are you installing the redis helm chart? Are you using the default parameters, or are you overriding some of them, like the password? And how are you injecting that information in the relay proxy helm chart?

[keelerm84]

Closing due to inactivity. Please feel free to re-open if you continue experiencing this issue.