Hi, thanks for reporting this, we should be able to fix this on our end. In the meantime as a workaround you may be able to set up your dagster daemon to automatically restart when it fails (with something like
restart: on-failure
in docker-compose, for example)