Space Reclamation Best Practices?

[Leezy 20]

what is the recommended Disk Reclamation schedule? how frequent do we run? i have near 1TB space to reclaim at offsite and we found out that each time we do space reclaim, the process will takes about 2-3 days. it is quite painful when we dealing with large data size. are there any specific recommendation to deal with this ? 

[Official Moderator 26]

Hello!

Please note that space reclaim maintenance is accumulative. It means you can stop it at any time and then continue from where you stopped it. 

Moreover, if the space reclaim is running but you need to run the job already, then space reclaim maintenance will be stopped automatically. 

These are the default settings. We start working with bigger data blocks and end with the smallest ones. That's why Space Reclaim maintenance is the most effective during the first 30-40% of the whole run. It's done in order to avoid the disk storage bottleneck.

 

[vger74656 1]

Would it be possible to schedule the stop of space reclaim?  That way it does the bigger blocks which accounts for the majority of the space.  Reason I ask is that we will have around 350 separate repositories and if the power goes out at any one of them, it corrupts the backups, and when this is multiplied by the amount of repositories, it makes it a nightmare to manage.

[Official Moderator 26]

It is possible to schedule only start of space reclaim.
In case Space reclaim maintenance would be run on weekly or two-weekly basis (depending on the used time, to minimize influence on production), most likely it will take much less time that running once per month, or once per two months.