John Davis
Regular Contributor
if you’re paranoid you can pull a drive ( degrading the array ), shove the new drive but NOT immediately rebuild onto it, instead initialise it as a new (jbod) volume to check it for errors and put some hours on it before committing to the rebuild
the bigger risk in yoru case is the worry that the weirdness you’re seeing with SMB is down to one of the existing drives being on the way out - the risk is when you pull a drive and rebuild the array onto a new drive the increased workload on the remaining old drives can push the failing drive over the edge. This is BAD if it happens - as in lose all you data.
So I’d definitely run a full backup of the array BEFORE doing anything else - either to an external or to another NAS
the bigger risk in yoru case is the worry that the weirdness you’re seeing with SMB is down to one of the existing drives being on the way out - the risk is when you pull a drive and rebuild the array onto a new drive the increased workload on the remaining old drives can push the failing drive over the edge. This is BAD if it happens - as in lose all you data.
So I’d definitely run a full backup of the array BEFORE doing anything else - either to an external or to another NAS