Is HA broken on master? about eirini-release

[srbry]

Hi @julz thanks for your response, I moved over to using kubecf and haven't had any issues, I can also confirm that just overriding the sizings for the NATs instance back to 1 did seem to work.

[cf-gitbot]

We have created an issue in Pivotal Tracker to manage this:

https://www.pivotaltracker.com/story/show/170951900

The labels on this github issue will be updated when the story is started.

[julz]

Hi @srbry - good news/bad news on this. Good news: we believe this is fixed in a newer version of scf, Bad news: unfortunately we're still using an older version, and may continue on an older version for a little while.

The context for the bad news is we're trying to avoid spending much effort keeping up with scf these days because we're planning - now that teams are spinning up to do this! - to start delegating the problem of deploying CF-with-eirini to other teams asap (e.g. via the new CF Operator from the Quarks team and the new cf-for-k8s effort).

Hopefully(?) for the moment you can work around this by staying with 1 instance of NATs and relying on being able to start a new instance quickly for HA. The bug should then be fixed either when we upgrade scf or - hopefully/likely sooner - when we move over to CF Operator/CF-for-k8s.