Configure health probes on Azure with Voyager

Hi,

I’m running an AKS cluster (Kubernetes v1.11) on Azure.

I noticed that when an Ingress is created, it will automatically create some entries in the cluster Load Balancer.

In my case, I have a few TCP services and I noticed that I get a lot of “spam” coming from Kube-Proxy due to Load Balancer Health probes. By default those checks are set to 5 seconds. To mitigate that problem, I increased the value from Azure Portal and my problem got solved. I noted that if the value is too big, Azure won’t route traffic over your ingress.

14%20PM

Does Voyager provides a way to set that Interval from a manifest/chart?

Thank you for your support,

Hey,

did you find a way? I stumbled across the same issue and try to figure a way out to fix it. As it is a ftps service it misbehaves if it is constantly checked by the external load balancer

I found no solution for that unfortunately. BUT, you really should read more about the difference between loadbalancers (standard and basic) here.

As you can read:

The minimum probe interval is 5 seconds and the minimum number of unhealthy responses is 2. The total duration of all intervals cannot exceed 120 seconds.

If you stay under the 2 minutes interval I think it could be a good compromise. On the Voyager side the Helm chart does not support that kind of configuration but we can see in the Azure documentation that we can customize this behavior using a Resource Manager Template. I think it’s possible to implement such functionality.

I hope it helped, have a great day!
Richer