Update on the EC2 outage

Our API is up and running again, and you can now use the CLI. A significant amount of deployments is still unavailable (they are on EBS volumes still affected by the outage).

We continue to work on the issue. Amazon did not provide an ETA to restore those volumes, and if the situation does not evolve, we will probably restore those volumes from snapshots and backups. If we have to do that, we will do our best to avoid unwanted side-effects. That means that if we restore from backups, we probably won’t put your service directly online, in case you want to restore from your own backups.

Important note: if your deployment is still down, you might be tempted to destroy and redeploy it. This won’t work, because that would mean that when the deployment comes back, you will end up with duplicate deployments. Our API won’t let you recreate such a deployment. However, you can deploy a new service with a different name.

Before trying to redeploy, look at your logs (with “dotcloud logs”). If you can check your logs, but your app doesn’t work, try a “dotcloud restart” – in some cases, that will do the trick.

We will keep you posted as the situation evolves.