This is for internal use by the PaaS team. Public-facing documentation is located at docs.cloud.service.gov.uk.

Restoring the Bosh Director

This guide will show you how to restore the bosh director and cells, in the event that it becomes unreachable or unresponsive.

Contents

Bootstrapping the director in another AZ

1. Using paas-bootstrap, bootstrap a new deployer-concourse instance using the BOSH_AZ env var to specify the target AZ. $ENV is the environment you’re targetting (e.g. dev or prod-lon).

BOSH_AZ=eu-west-2b make $ENV deployer-concourse bootstrap

1. Wait for the bootstrapper VM to come online, and access it at localhost:8080
2. Run the create-bosh-concourse pipeline from the left-hand edge

The director’s database is gone

In the event that the director’s database is gone, it can be restored from the latest snapshot to get to back to a state from which the rest of the deployment can be fixed.

1. Find the latest snapshot of the director’s database in the RDS console
2. Create a new database from this snapshot, using exactly the same name as set by Terraform in paas-bootstrap/terraform/bosh/rds.tf
3. Bootstrap a new deployer-concourse using paas-bootstrap in the same region.
4. Wait for the new database to be available.
5. Run the create-bosh-concourse pipeline from the left-hand edge. During the bosh-terraform step, Terraform will find the same-named database and configure its parameter group, security group etc correctly.
6. The director will be deployed with access to the correct database.