Evidently, availability of a storage system is not that great if it consists of disks in RAID-5 configuration. I am not operating RAIDs myself, but from what I read already some years ago, mirrored schemes are the way to go nowadays if availability is desired (and disks required).
/armchair storage expert mode off
Probably has a lot more to do with limited budget than anything else. They're still using a system with HDDs instead of SSDs because of budget reasons.
But the main issue with RAID5 and replacing a bad drive on it is the rebuilding of the array with the new drive. If their was little to no changes on the disk array while it rebuilds it would go quickly, but since the database is constantly being changed/updated while it rebuilds it slows down the rebuilding process by a lot.
I recommended them on their forum to just shut the project down for however long it takes to rebuild the new drive. I'm guessing that's what they are doing now.
Also, their could be other issues going on that they haven't been able to figure out due to the bad drive. Whatever that issue is could have led to killing the drive also.
Facebook
Twitter