Open Racer159 opened 6 months ago
High level: Follow practices here for services. Keep everything possible in aws provided storage (RDS, S3, Elasticache) Ideally manage secrets externally to mitigate risk and ease restore process. Implement functional equivalent or linux ec2 backup/restore steps on the k8s pods that they are analogous to.
@oates add spike label
We should sync with @corang as well since his team has started down this path already
This may also help inform what we want to say for this issue: https://github.com/defenseunicorns/uds-common/issues/130
Is your feature request related to a problem? Please describe.
As Ashton I want to be able to backup GitLab safely so that I can restore it easily in the event of a disaster.
Describe the solution you'd like
We should identify and select solutions that are able to backup the variety of places that GitLab stores data:
https://docs.gitlab.com/ee/administration/backup_restore/backup_gitlab.html
And then follows the recommendations for restoring a GitLab server:
https://docs.gitlab.com/ee/administration/backup_restore/restore_gitlab.html
These should be written in a simple ADR to capture why the final solution was chosen and a follow on implementation issue should be created.
Additional context
We currently don't have set SLAs but should target a baseline to have something while we continue to experiment.