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.
Bootstrapping the director in another AZ
paas-bootstrap, bootstrap a new deployer-concourse instance using the
BOSH_AZenv var to specify the target AZ.
$ENVis the environment you’re targetting (e.g. dev or prod-lon).
BOSH_AZ=eu-west-2b make $ENV deployer-concourse bootstrap
- Wait for the bootstrapper VM to come online, and access it at
- Run the
create-bosh-concoursepipeline 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.
- Find the latest snapshot of the director’s database in the RDS console
- Create a new database from this snapshot, using exactly the same name as set by Terraform in
- Bootstrap a new deployer-concourse using
paas-bootstrapin the same region.
- Wait for the new database to be available.
- Run the
create-bosh-concoursepipeline from the left-hand edge. During the
bosh-terraformstep, Terraform will find the same-named database and configure its parameter group, security group etc correctly.
- The director will be deployed with access to the correct database.