Hello

 

I currently have a 1 TB nvme drive as my cache, however I've bought 2 new drives (also 1 TB) to replace as a cache pool in a raid 1 and then use the old drive just for caching share data / new downloads.

 

My plex metadata folder is large with a ton of files, so I am trying to find an easy way to do this. I wanted to confirm if my strategy would work, and if there are any potential issues.

 

1) Add one of the new 1 TB drives to the existing cache pool (which I assume will start to mirror the existing one). 

2) Once replicated, remove the old cache drive, and replace it with the 2nd new 1 TB drive and let the mirror rebuild

3) Re-add the old cache drive in a new pool for downloads

 

So, a couple questions:

1) Am I correct in how the RAID 1 system will work? Do I need to do anything to make this happen?

2) Is there a way to know once the 2 drives are completely mirrored

3) Are there any issues with the above approach

 

That works assuming current cache is btrfs, you'll know when the btrfs operations are done when the stop array button is available again.

4 minutes ago, JorgeB said:

That works assuming current cache is btrfs, you'll know when the btrfs operations are done when the stop array button is available again.

 

The current cache drive is btrfs, so I should be good.

 

Thank you for confirming.

What if you have 2 SSD drives as cached pool I want to do something similar but add NVME drive so can I remove one of the BTRFS SSD to simulate a failure then add the NVME as a drive to rebuild to then remove the other SSD after the NVME has finished its rebuild and add a further NVME to rebuild from the new NVME to second NVME having a pool of 2 x NVME Cache pools.

If I understand correctly it's possible, but safer to add the new device(s) and remove the old one(s).