OCZ Vertex 3 Fail: “Runtime corruption detected on SSD” and why SSD drives are not safe yet.

After almost 9 months with the OCZ Vertex 3 SATA-6GBPs (read about my OptiBay two-drive setup) and mostly no issues, the drive died on me. Starting with several GSOD (Grey Screens of Death) on my OSX Lion MacBook Pro, the drive became more and more unreliable.

Nov 28 18:30:42 mbp kernel[0]: hfs: FindNextLeafNode: Error from hfs_swap_BTNode (node 9852)
Nov 28 18:30:42 mbp kernel[0]: hfs_swap_HFSPlusBTInternalNode: catalog key #54 too big
Nov 28 18:30:42 mbp kernel[0]: hfs: Runtime corruption detected on SSD, fsck will be forced on next mount.
Nov 28 18:30:42 mbp kernel[0]: hfs: FindNextLeafNode: Error from hfs_swap_BTNode (node 10370)
Nov 28 18:30:42 mbp kernel[0]: hfs_swap_HFSPlusBTInternalNode: catalog record #22 keyLength=32 expected=65568
Nov 28 18:30:42 mbp kernel[0]: hfs: node=10965 fileID=4 volume=SSD device=/dev/disk0s2
Nov 28 18:30:43 mbp kernel[0]: hfs: Runtime corruption detected on SSD, fsck will be forced on next mount.
Nov 28 18:30:43 mbp kernel[0]: hfs: FindNextLeafNode: Error from hfs_swap_BTNode (node 14503)
Nov 28 18:30:43 mbp kernel[0]: hfs_swap_BTNode: record #55 invalid offset (0x9B46)

It was getting slower and slower, reporting tons of errors in the Console.app. Then, it did not reboot anymore. Even in an external USB drive enclosure the drive did not mount anymore. Contacting OCZ about this issue, they sent me an RMA number right away and told me they will replace the drive. Does this mean our data is not safe in SSD? What can we learn from this?

First of all, it is a myth that SSDs are more reliable than hard drives. Secondly, they are not even fully supported on OSX yet (see Should I use Trim Enabler on Lion for the OCZ Vertex 3? No!) - at least if you install the drives yourself (which is also what I would suggest, since Apple does not offer any decent SATA 6GBPs drives yet).

So where is the proof, you ask?

Proof #1: The Hot/Crazy Solid State Drive Scale (via Coding Horror)

Super Talent 32 GB SSD, failed after 137 days
OCZ Vertex 1 250 GB SSD, failed after 512 days
G.Skill 64 GB SSD, failed after 251 days
G.Skill 64 GB SSD, failed after 276 days
Crucial 64 GB SSD, failed after 350 days
OCZ Agility 60 GB SSD, failed after 72 days
Intel X25-M 80 GB SSD, failed after 15 days
Intel X25-M 80 GB SSD, failed after 206 days

As a commenter put it: "Average life of SSD = 227.375 days (based on Wills' data)" - which is also what I can confirm.

SOLUTION: Backup early, backup often. Don't rely on the SSD and make two local backups plus one backup in the cloud.

Proof #2: long-term study of SSD failure rates (via Tomshardware)

SOLUTION: Buy drives that come with a very long warranty. Be prepared to let your drive replace for several times.

I am still waiting for my replacement OCZ Vertex and I will benchmark how fast it will die again. Since OCZ told me, it can take up to three weeks for my replacement to arrive, I bought myself a Seagate ST750LX003 750 GB SATA 600, Momentus XT, 8 GB SLC - it's a hybrid 750GB HDD with an 8GB SLC cache. The drive shifts data that is used often in the 8GB SLC SSD space automatically. While the SSD part of this hybrid drive can also fail, the data won't be lost and it will suffer just some minor speed loss (if it fails, since SLC is said to be more reliable than MLC chips). While this disk is now my main startup disk, i will go back to using the OCZ Vertex 3 again, once the replacement arrives. I just need to backup regularly - the speed gain is really worth the trouble.