[gogeek]

We had about 40 OCZ drives (Vertex 2/3/3 Max IOPS) and 6 of them failed. Our 8 Vertex 3 Max IOPS were in RAID 5 for a huge calculations which didn't require reliability but wrote a lot of data and therfore we tried to save time with this experimental RAID. The RAID was fine for about a couple of month and seeing almost 3 GB/s throughput was mindblowing. But suddenly we we saw drives randomly failing. But the drives did not completly fail, we were able to rebuild the RAID with the same hard drive. We did that a couple of times until we thought it was too much hassle and used the drives somewhere else. Now we are buying Crucial M4s and they are totally fine. In the first place a hard drive must be reliable.

[danielvinson]

This was a firmware bug originating from the Sandforce firmware which the releases were based on - I think this one was completely fixed at least 8 months ago.

[arprocter]

We had to replace 50+ Agility 128s as around half of them went bad within weeks of install.

A year later and none of the replacement Intels have gone bad...