Clusters, custom health check paths, deploy commit SHAs & a redesigned dashboard
One Phoenix app across several servers, plus a fresh look for the whole app.
Clusters
You can now run one Phoenix app across several app servers at once, behind a load balancer, with the database on its own server. If a node goes away, the rest keep serving.
Attach, promote, and detach nodes from your app's new Cluster page. Deploys roll node by node, with each one health-checked before traffic switches, and migrations run exactly once on the primary. Apps that rely on Phoenix.PubSub or Presence across nodes can turn on BEAM clustering with one click.
Get started with the Clusters and BEAM Clustering docs.
Custom health check path
You can now choose which path Potions requests to decide your app is up. Every app is checked before traffic switches to a new release, and every minute if uptime monitoring is on. Until now that check always hit /. Set a dedicated route instead, such as /health.
See the Health Checks docs for more.
Deploy commit SHA
Every deploy now injects POTIONS_COMMIT_SHA into your app's environment, carrying the full commit SHA of the release that's running. Point AppSignal's revision, Sentry's release, or your error tracker's equivalent at it, and errors group by deploy instead of by wall clock. After a rollback, it reports the SHA that was actually restored.
See Deploy Metadata for setup with AppSignal, Sentry, Honeybadger, and Bugsnag.
A redesigned dashboard and app
The dashboard is now a fleet view, which scales as more servers and apps are added. We've also updated the design of all app pages.
Show improvements Hide improvements
- Caddy now restarts itself automatically if it fails, on new and existing servers
- Fixed: creating your first DigitalOcean server in a region your account had never used failed with a 403
- Deploy logs show the HTTP status for every health check attempt
- Health check failure emails name the URL that failed