I have a raid5 enclosure with 5 Fujitsu F1 750gig sata drives, and I have one drive that keeps dropping out. I ran MHDD on it and did a scan, and it got the following scores: 5249631 <3 ms, 494458 < 10ms, 1593 < 50ms, and 2 < 150ms. I don't see anything that looks like actual errors. After looking around, I noticed that Western Digital makes a line of sata drives that are raid specific and it says specifically that they do something different when errors occur to avoid drives dropping out, so it sounds like this is a common or at least known issue. So I guess I screwed up when I bought these Fujitsus, but they were very highly rated for a good price, and I didn't know any better. My understanding is limited, but the way I read it, when a HD encounters an error, IDE drives try to handle the error internally, which can cause a delay, and if the delay is too great, then the drive drops out of the array. If the numbers above sound like the hdd is actually bad, I can RMA it, but I really need to get this drive back in to restore fault tolerance asap, as I have 3tb running on basically a 4-drive raid 0 at the moment, ie bad news. Any idea what kind of delay numbers would cause the drive to drop out, or does anyone have any advice for me at all? I actually bought another 750 for other purposes, but it was purchased separately, and it seems to be a very slightly different internal version number or something, because the array rebuild process fails with that drive too. I've been trying to rebuild the array for 2 weeks now (rebuild takes 26 hours, so I get to try once per day). Basically the rebuild runs, and about halfway into it, this drive fails, the raid enclosure reports the bay as "empty", and the rebuild stops. This happens over and over. I would love to run the mhdd benchmark against one of the functional drives for comparison, but I'm skeered to pull them at this point. I have new drives coming, and I'm going to back up the whole thing. Since I'm already in mhdd, I'm going to go ahead and try to low level format it and then put it back in the array and run rebuild again, but I'm really just guessing here.

In case it matters, the message in the raid utility says "wait for reset notification", then the drive id, then "ProcessDroneSrbExtensions"

Much appreciation in advance for any and all advice.

Here is a screen shot from the raid management utility

Sorry, I'm an rtard, it's a Samsung, not a Fujitsu.

copy the information somewhere safe, check each individual disk a replace the bad ones, check your controller, and re-build the raid

oh yes and raid specific disks have timing to allow use in raids, desktop disks don.t. REx (from Western) or Ultrastar from Hitachi are some that do