Disaster Recovery and Passive Standby Config for Hasura Cloud
Introduction
If you're designing your system to be fault tolerant enough to account for a full region-wide outage, we recommend enabling a backup region on Hasura Cloud too.
Step 1: Create a secondary Hasura Cloud Project
In addition to your primary region on Hasura Cloud, create a new project in a convenient secondary region of your choice and within a new VPC - we will call this the secondary Hasura Cloud project.
This guide assumes that you use a custom domain for the primary Hasura Project so that in the event of a failover you can switch the DNS for the custom domain to point to the secondary Hasura Project.
Step 2: Peer the VPC and configure
Peer this new VPC with your disaster recovery region on AWS or GCP where your secondary database is present.
Create all the required configurations and environment variables for the secondary Hasura Project. Connect this project with your secondary database.
Step 3: Keep Metadata in-sync
Depending on when and how you keep the secondary database schema and data synced with the primary database, you should also keep the Hasura Metadata on the secondary project up-to-date and in-sync with the primary project. You can use our Github Integration features to keep your secondary project in-sync by deploying each commit to both projects. Care should be taken in the config of databases and ENV variables in this regard.
Step 4: Redirect traffic on failover
When you actually need to failover to the secondary database and the secondary Hasura Project, point your custom domain to the secondary project by changing the DNS entry on your side. This will send all traffic through to the secondary project into your secondary database.
Scheduled events (which are not part of Metadata), pending events deliveries, as well as the events history, will not be migrated in the event of a failover.
