GCE: Converting ephemeral IP to static doesn't update related TargetHTTPSProxy

[kimroen]

This could be user error or issues with documentation, but I was encouraged to make an issue after first commenting on the help document I was following, so here goes:

I was going through this tutorial on getting an Ingress set up, and then later ran into some issues getting TSL-termination working. When visiting the domain using https://, the browser didn't get a response from the server.

After some head-scratching, I realized the problem I was having was that the TargetHTTPSProxy and TargetHTTProxy had different IP-addresses, so HTTPS-traffic never reached my URL map.

Step 5 of the document linked above explains how to convert the ephemeral IP you get when creating the Ingress to a static one. I had followed this advice, but only the TargetHTTPProxy was using this static IP, and the TargetHTTPSProxy was still using an ephemeral IP that had changed.

Some speculation on a few different things the problem could be:

1. The TargetHTTPSProxy didn't use the static IP automatically as I was expecting
2. The kubernetes.io/ingress.global-static-ip-name annotation on the Ingress needs to be set when converting the IP
3. HTTP and HTTPS were already using different IP-addresses before I converted and I just didn't notice I was only converting one of them

Hopefully that helps!

[nicksardo]

Which version of Kubernetes were you using?

[kimroen]

This is a newly created cluster, so it's running the default (1.8.8-gke.0).

[fejta-bot]

Issues go stale after 90d of inactivity.
Mark the issue as fresh with /remove-lifecycle stale.
Stale issues rot after an additional 30d of inactivity and eventually close.

If this issue is safe to close now please do so with /close.

Send feedback to sig-testing, kubernetes/test-infra and/or fejta.
/lifecycle stale

[rramkumar1]

/remove-lifecycle stale
/lifecycle frozen

[rramkumar1]

@kimroen Is this still an issue?

I'm curious how this happened because if you create an HTTP(S) Ingress from scratch, we automatically allocate a reserved IP for you.

[kimroen]

@rramkumar1 Hi! I fixed it manually when it happened so I'm not sure if it's still an issue with new Ingress setups.

Like I say in the original issue, this happened as I was following a tutorial (linked above), so it's very possible that I messed something up or that the tutorial was unclear. In either case, I ended up with two proxies using different IP-addresses, one ephemeral and one static.

[swetharepakula]

Closing since there has been no update since 2019.