I installed a 4x8TB SATA disk (+1xSSD per system) on my home NAS (Supermicro X7SPA-HF)
BTRFS volume, RAID 10, one big 16TB pool and Rockstor 4.1.0.0
Now one of the drives is faulty and I have not found ANY simple procedure to replace that drive so that rocktor can handle the exchange and data recovery ideally with a “one click” solution. In most procedures, backup the data, break the pool and rebuild, restore the share, and load the data from the backup. 16TB of data is really not easy and certainly not cheap to just back it up somewhere, because the system is not able to easily deal with such a basic issue as one defective disk in a raid.
So how do you do it WITHOUT shutting down the NAS, without losing data, and ideally without breaking access to services like NFS and Samba?
Is it possible to put buttons or some menu item in future versions of rockstore that would solve this easily? When I disconnect the faulty disk while it is running, it is deattached, the new one will not take it, resize/repool and deleting the disk does not work, because raid 10 needs 4 disks, etc.
You can also that your feature request is already noted, but not implemented yet, due to the ongoing overhaul of Rockstor (moving off obsolete python versions, upgrading the django framework and the like, that’s driven by @phillxnet and @Flox).
We also had a recent thread around this (replace vs. add/remove), which you can find here:
I used the add/remove disk through the WebUI, but that took quite a bit of time, since I replaced all of my disks in the RAID pool.
either of these options should allow you to perform the actions without shutting down the NAS. In my opinion though, I would avoid major write operations during that time (e.g. backing up massive amounts of data to the NAS from other sources), as that will slow down the replacement process.