OIM require database snapshots to be placed in a volume called "archive". These snapshots shall be created automatically on a schedule.
Feature
Create a persistent volume claim - confirm with OIM naming ("archive"?) and location (one shared for all workloads: sdis, wastd, tsc, etc. or one for each workload?) of a size within the k3s cluster's allocation limits but large enough to hold db dumps
Mount volume to SDIS PROD, SDIS UAT, and a new container running the archival cronjobs
Create a container with
a daily cronjob running a script to pg_dump PROD db to file sdis.dump, then pg_restore UAT db from file sdis.dump
a weekly cronjob to pg_dump PROD to a dated snapshot, e.g. 2022-01-01_sdis.dump
somehow prune old snapshots (ideally when they've been harvested for long term archival by OIM)
Questions to OIM
Which namespace for the volume: default or per app (sdis, wastd, tsc, etc.)
What name for the volume: "archive"?
Any preferences for db snapshot frequency and naming scheme?
Problem
OIM require database snapshots to be placed in a volume called "archive". These snapshots shall be created automatically on a schedule.
Feature
Create a container with
sdis.dump
, then pg_restore UAT db from filesdis.dump
2022-01-01_sdis.dump
Questions to OIM