Closed esune closed 1 year ago
It's a good practice to keep production data backed up off-site, or to multiple destinations.
CrunchyDB's pgBackRest support configuring multiple repos
for backup storage.
By default, a PVC will be used to store the backup in dev
, test
, and prod
.
@WadeBarnes Should we setup prod
with backups going to AWS S3 or BCGov's Dell EMC appliance as well?
I'm assuming the backups will be saved to their own PVC. The PVC should be provisioned as netapp-file-backup
which is described here; https://github.com/BCDevOps/backup-container#backup-storage-volume, which writes to off site (cluster) backup storage automatically. Platform services will soon be recommending S3 storage for backup retention. This will be supported by BC Gov S3 buckets. @jleach should be able to provide additional details regarding the move to S3 storage.
For the backups, is there an option for notification of success/failure? The backup-container notifies us of all backup success/failure in the #ditp-backups
channel. I'd like to see the same for the CrunchyDB backups.
I can look into it.
traction-database
CrunchyDB HA cluster in dev
7
days worth of continuous backup (point-in-time restore possible)netapp-file-backup
storage classdev
has been migrated to use traction-database
clusterExcellent work @i5okie . We need to plan for the upgrade of test
now. In the meantime, would you mind deploying the production instance (#132), since we are not worries of losing data there?
Traction prod
has been deployed on October 23rd.
Traction test
has been upgraded to use CrunchyDB.
PR with updates to the deployment values:
https://github.com/bcgov/trust-over-ip-configurations/pull/171
Upgrade dev/test deployments of Traction to use CrunchyDB rather than standalone PostgreSQL as database provider.
Pre-Requisites:
Requirements:
dev
test