We plan to switch from Nexus version 2.12 to version 3.2.
What can you recommend for a large volume backup (1-3 TB) of BLOB storage in Nexus 3.2?
1. Configure the Export configuration & metadata for backup task to export databases.
2. Run the scheduled task to export the databases to the configured folder.
3. Back up custom configurations in your installation and data directories around the same time you run the export task.
4. Back up all blob stores.
I think that the implementation of 1 to 3 steps of the common scenario will not take long. Suspends access to the database until the backup is complete.
But the execution of 4 step can last for more than an hour, even if we talk about incremental copying. During this time, access to the database will be restored.
What will happen if, during the backup of the BLOB storage, the artifact will be deleted from the repository. The database will delete the entry, the group of blob files will be deleted in the BLOB store, but the backup will continue to refer to the file from the BLOB of the repository, since the backup was created relatively long ago.
When restoring a backup, how does this collision manifest itself?
Or is this situation not critical for such large volumes?
Can I first make a backup copy of the BLOB repository, and then run the scheduled task to export the databases to the configured folder?
Regards,
Anatoly Laskin