My team has implemented a custom backup process for Bitbucket Server. On a nightly basis, we take down Bitbucket, rsync the files, take the DB backup, and save everything to tape. This covers the major disaster recovery sceanrios by allowing the entire instance to be restored as of the prior night immediately and older instances via tape recovery.
In addition to this, my team would like to implement nightly backups at the repo level as well. This would cover the case of a single repo being deleted. While a full nightly restore is possible but may not be warranted if a repo could be created and cloned into Bitbucket. This would result in less disruption to other teams. We plan to keep only a few days rolling backups. As a sanity check, does this sound reasonable?
If so, how would I go about to iteratively clone all repos within a Bitbucket instance? I need it to be adaptive since I do not wish to update the script for every newly created repo.
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.