Integrating backups into your GitOps Pipeline

What if a single bad commit could destroy your database? Learn how to integrate automated snapshots into your GitOps pipeline, creating a restore point before every deployment.

#1about 6 minutes

Why redeploying code is not a substitute for data backup

GitOps pipelines excel at redeploying stateless applications but fail to protect persistent data in databases and volumes which are not managed by version control.

#2about 2 minutes

Demo overview of the GitOps backup and recovery workflow

The demonstration shows how to use a pre-sync job in Argo CD to automatically create a data snapshot before applying a potentially destructive code change.

#3about 5 minutes

Deploying a stateful application with an initial backup

A stateful application is deployed via Argo CD, and a pre-sync hook automatically triggers an initial backup job to establish a clean restore point.

#4about 4 minutes

Simulating data corruption from a bad code commit

A bad commit with a faulty SQL statement is pushed, but the CI/CD pipeline first creates a snapshot before applying the change, which then visibly corrupts the application's data.

#5about 4 minutes

Restoring data and applying the corrected code change

The application is restored to its pre-failure state from a snapshot, after which a corrected code commit is successfully deployed through the same automated backup-then-sync pipeline.

#6about 2 minutes

Key takeaways and resources for learning more

CI/CD pipelines are essential for stateless applications, but protecting stateful data requires integrating backup tools via APIs, with resources like Project Pace and CubeCampus available for hands-on learning.

#7about 3 minutes

Addressing backup efficiency for large data volumes

Using storage-level snapshots for backups is highly efficient as it is nearly instantaneous regardless of volume size, unlike a full data export which would be time-consuming.