Tero Marttila

> What happens when affinity can not be satisfied? 1.3.4 says RuntimeError IIRC That's an instance deploy error. Those should be reported the same way under both 1.3 and 1.4:...

> yes, RuntimeError can not be right. Fixing that in #2931... not directly part of this issue.

The specs and e2e specs are now green for the this entire series of PRs, so this should be feature-complete.

The `Kontena::Launchers::Weave`, `Kontena::NetworkAdapters::Weave` and `Kontena::Workers::WeaveWorker` should now be in shape, the specs pass, and no known issues in testing. Now need to refactor the `Observer` -> `Observable` update-chaining.

Note: I don't really expect this PR to get reviewed/merged in this form. This is more about experimenting with how I think the agent could get refactored... once I think...

FWIW upgrading terraform and the DNS provider makes no difference, I can still repro the `NXDOMAIN`s: ``` $ terraform version Terraform v0.12.16 + provider.dns v2.2.0 + provider.hcloud v1.11.0 + provider.random...

It looks like fixing this as-is would break the use of authenticated [docker registry proxies](https://docs.docker.com/registry/recipes/mirror/) with `registry-mirrors`. The only way to use these currently seems to be: * Register a...

We ran into the same issue as a regression, where cluster metrics stopped working after an upgrade from 13.1.0 -> 14.0.1. Perhaps this issue should be re-opened, as #464 has...

Only adjusting the `https://` protocol part based on the HTTP X-Forwarded-Proto header is not enough, as the `$_SERVER['SERVER_PORT']` will remain 80, leading to URLS of the form `https://host:80/` when behind...

https://github.com/terraform-providers/terraform-provider-aws/pull/12574#issuecomment-639642524 explains that this is an API issue with `type = "forward"` + `target_group_arn = ...` vs multiple `target_group { arn = ... }` blocks... > Hi @jmgreg31, I think...