Question: How does the restore handle duplicate data?

[andrewfenn]

Hello,

Sorry to bother I wanted to ask about something not in the documentation. How does the restore handle duplicate entires?

I have lots of extremely large DBs that I want to backup and move but I don't have enough time of a maintenance window to ensure all data is captured before switching. My plan is to:

1. Backup the data and restore on new system
2. Switch everything to point to the new system
3. Backup the old DB again and restore on the new system again to add any additional records they may have got created while steps 1 and 2 were working.

Would it be safe and viable to do? Do any duplicates on the second restore at step 3 get ignored while new entries that might have been added while in the maintenance period are still added to the new DB?

[dalgibbard]

Its been a long time since I used this, but iirc, when trying to restore an object which already exists, it throws a 409 conflict error, and doesn't apply.

I don't recall if this breaks the whole restore process or not though.

Edit: looks like it captures curl errors with retries using $attemptcount. I'm not sure if curl throws a non-zero error for 409 conflicts, but if it does, you could comment out those retry/fail sections as an option, or amend it to exclude a particular error code.

Edit2: see also: #79 (comment)