Aws-load-balancer-controller: Does aws-alb-ingress-controller support setting health check timeouts?

Created on 22 Nov 2018 · 6Comments · Source: kubernetes-sigs/aws-load-balancer-controller

I have been trying to set health check related timeouts in my ingress configuration:

    alb.ingress.kubernetes.io/healthcheck-interval-seconds: 5
    alb.ingress.kubernetes.io/healthcheck-timeout-seconds: 2
    alb.ingress.kubernetes.io/healthy-threshold-count: 5
    alb.ingress.kubernetes.io/unhealthy-threshold-count: 2

But when ALB is created it doesnt seem to override any of the above timeouts?

I'm on v1.0.0 of aws-alb-ingress-controller. I was following logs and even there I didnt see any of those timeout config while target group creation

Source

drop-harsh

👍1

Most helpful comment

I guess the issue is with passing them as integers. If we pass them as strings it works. Looks like it was documented incorrectly.

kumud-pe on 23 Nov 2018

👍3

All 6 comments

I have a similar issue.

Whenever I use above annotations the alb-ingress-controller just ignores the ingress and is not creating proper AWS Load Balancer and Target Group.

Also, setting these annotations on the Service does not take any effect.

krzyho on 23 Nov 2018

👍1

Facing the same issue. If I use these annotations ALB is not getting created at all.

kumud-pe on 23 Nov 2018

I guess the issue is with passing them as integers. If we pass them as strings it works. Looks like it was documented incorrectly.

kumud-pe on 23 Nov 2018

👍3

Yeah passing them as strings seems to work

drop-harsh on 23 Nov 2018

Sorry for the late reply.

Yes, these annotations can only be strings. and boolean/integers must be quoted(otherwise, other annotations will be lost as well, i think kubectl should be improved to give error or warning for such case).

I'll update the docs today, sorry for the trouble caused