Question on Pod Readiness gate & Zero-Downtime Rollouts on AWS ALB

[deybhayden]

Hey all, I've been reading the documentation on Zero-Downtime rollouts on AWS ALB.

I have some logistics questions on enabling the target group verification and/or pingPong verification.

1. Are the two paths I mentioned exclusive? If I did one, I wouldn't do the other?
2. Do I need to set up the ALB controller label elbv2.k8s.aws/pod-readiness-gate-inject: enabled in the pod namespace per the ALBC documentation, or do the two paths I mentioned above not require this?

Sorry for the confusion, just trying to sort my next steps on how to enable this in my cluster. I currently have simple canary rollouts on ALB operating in IP mode, but am experiencing downtime when the IPs switch when my new pods come up. Thanks!

[deybhayden]

Reading on this issue:

The v1.1 Target Group verification feature was implemented to provide zero-downtime guarantees in the absence of proper readiness gate injection, and it should satisfy any concerns of zero-downtime. But in case the feature is not used, Argo Rollouts could offer a model where selectors of services are modified in a way that works well with ALB LoadBalancer Controller's implementation of readiness gate injection.

Leads me to believe if I want to let Argo Handle the validation, I can use the Target Group Verification feature - without setting up the label on the namespace. I'm going to head that direction and see where I end up.