AbsaOSS
Execute Enceladus application data backup and restore
Background [Optional]
A clear explanation of the reason for raising the question. This gives us a better understanding of your use cases and how we might accommodate them.
Goald
TBD
The actual BU&R activity depends on how backup environments are managed. Two potential options spring to mind:
- a new DB running alongside the existing DB, using a backup of the existing DB
- an isolated environment, also using a backup of the existing DB
Procedure in either case:
- Select one Menas environment from which to test BU&R (in the case of 3 machines, for example, the third)
- Limit access to those doing the BU&R test so potential users are not inadvertently accessing the instance connected to the backup DB
- Point the Menas env at the backup DB
- Update Enceladus configs (perhaps in an isolated env with fresh configs) to point to the selected Menas environment
- Verify data is as expected on the Menas env
- Perhaps run a test job from Enceladus to verify no issues with the backup DB, etc
After testing is concluded:
- Point Menas at the original DB
- Return normal access for users
- Revert Enceladus configs if changed
This is more a test of the backup rather than Menas/Enceladus, as nothing major is changed there. Verification of the data is the goal - be it count of items, data integrity, normal functioning of the DB, etc.
We need to be clear on how we are validating the data is landing as expected for example are we making a copy of last night's data that landed and today we spin up the backup and restore and run that same pipeline and see if the data matches ? We need data test proof for this task
I am missing for how long the backups have to be stored. Or is that about keep only the latest one?
I have added pre- and post- counting/hash script results in the section Recovery Point Validation.
