In RHODS-1977, a bug in HA behaviour has been fixed and now users are able to access their JL servers even if the JH leader pod dies while they are in the spawning process.

However, if the leader pod dies while the user is already being redirected to their server, they will see a Gateway Timeout error page.

While this is a low probability event, a possible fix for this corner case should still be investigated; the UX is not great in this case, and perhaps even a custom error page asking the user to reload the page (thus being redirected to their server) would be enough to make it acceptable.