Sometimes, when reusing a Brancher older than a week, there is no ensure_copied_app / ensure_app flow in the logbook anymore. So it will 404 all the time, until it reaches the timeout, then it raises TimeoutException.
When reusing a Brancher, we should try to poll for SSH first, then if it isn't reachable, do the wait for availability process.
Example job: https://github.com/ByteInternet/hypernode-docs-next/actions/runs/20754105263/job/59658811645