Ian Fleming used to have James Bond say, “Once is happenstance, twice is coincidence, but three times is enemy action.”
I had my third Seagate Barracuda 3TB drive fail in a little over two years. Then, as the RAID was rebuilding, I had a fourth Barracuda bite its last tuna.
Luckily, the RAID level was 6, so I can tolerate two simultaneous failures without data loss.
The day started out bad when I heard the Synology NAS box beeping piteously. A look at the GUI showed what was wrong. I put a 4 TB Seagate drive in and started the rebuild, but soon noticed another failure:
Disk 4 is rebuilding, but disk 8 is MIA. You’ll notice a lot of 4 TB drives. Each one is there because a 3 TB Barracuda failed.
After the rebuild finished, I installed a new disk in position 8 and started that rebuild.
Why are the disks failing? Not because they’re hot:
To save you the calculating, 19622 hours is 2.24 years. Too soon, if you ask me.
The number of disk suppliers is waning rapidly. Western Digital bought the Hitachi disk business, and then Seagate bought WD. I’ve always liked the Hitachi/IBM drives, but they’re hard to find and expensive now. And who knows when somebody is going to save a few bucks and build ’em like the Barracudas.
Still, it’s cheaper to by consumer level disks and replace them when they fail then buy enterprise drives. Until you lose data. If I were you, I’d forget RAID 5.
Leave a Reply