identical nlbs, one has intermittent health check failures

100% it was a flakey nlb, an issue on the AWS service side.

Both my nlbs are set up with cloudformation, i tore them down, recreated them and the issue went away.

one of the elastic ips i associate in the cloudformation to the nlb took a good few minutes to disassociate, showing errors when i tried to manually delete, but eventually resolved itself, and after that i was able to spin up the nlbs with my cloudformation again and have not seen errors since. i had a small panic about that elastic ip tho, this is wired up to lots of customer systems, if that were broken i'd have been emailing a few hundred people to try to move them.

what is frustrating is that there's no support path for issues like this, this is the second time i've run into an error that is certainly and entirely on the aws side, but how do i get support on failing services that I pay for? Fixing this meant tearing down a production service and going from bad quality to no quality for a few minutes for my customers, is there no way to get support for legitimate AWS bugs? I tried the feedback link in the aws web console and even supplied the arns for my nlbs, but that also seems to be just shouting into the void. it's very frustrating.