PG Datastore Utilisation-Phase 2

~150 hours per month

~40 bridges this month

~200 alerts per month

Example: https://jira.bmc.com/browse/SSR-11131

BHOM_VC :ams3-vcnt-ctrl3.onbmc.com@datastore-1368088_AMS3-PGDB-00:01:101 disk usage is > 75.0 75.800476 reported from ams3-vcnt-ctrl3.onbmc.com

1. Identify which PG servers are on tha datastore.

2. Identify how much space they are using.

3. Identify which PG servers can be migrated (not leaders).

4. Migrate 1 or more databases

Phase-1 (Delivered)

1. Need read only access to vcenter ( we need a service account)

2. Search for a specific datastore

3. Go to VM list- We can see VM details with usage.

4. In Summary Tab we can see total usage details etc.

Phase-2 (In Progress)

1. Leader can be checked in postgres cluster or DB server by DB team ( use putty to check patroni cluster to check details)

2. To migrate, we have to bring the VM down and click on Migrate to a different datastore (Downtime is needed)

3. Migration of replica server is possible in production only.

1. Find the VM using the most space (but less than 1TB*) from phase-1,

2. Find the datastore on the cluster that will have more than 25% free AFTER migration,

3. Write the proposed migration to the incident ticket.

*If the VMs are using more than 1TB, Broadcom recommends stopping them before migration.

Phase-3 (Future)

1. Perform the migration recommended in phase-2.