Short answer: Test the bad drive of a RAID array to see if it's really bad. If the status happens to be a false negative, erase the drive's first 10 sectors, and let the RAID array rebuild itself.
*****
I added a new storage peripheral to my system a couple of weeks ago, but I just got it to work these last few days. It's a USB3 drive enclosure featuring RAID configurations. It's not directly related to FreeBSD, but I do use the UFS file system with it, and the tips I learned could help a fellow FreeBeastie. It looks like these kind of units are sold under different brands, but many seem to use a JMICRON TECHNOLOGY RAID controller. Mine is the Vantec NexStar HX4R.
While trying to install it, maybe I did something wrong according to Murphey's Law, I ended with a bad drive according to the flashing blue LED of the unit. Moving the hard disk into a different slot made it clear that the problem was with the hard disk, Page 0 Damage Error!
I had doubt about the hard disk being bad because I had it for about 6 months, and never really used it much. How disappointing! A quick read test using the
(I use daX throughout this post to prevent disasters in case the ENTER key is pressed before changing the device name. Btw, daX is a 3Tb drive.)
Filled with doubts because a sample test successfully passed doesn't prove the absence of a problem, I continued with the installation of the hard disk manufacturer's full test application. It too didn't find any problem. Me very puzzled, so me do more Internet search digging... Assembling several pieces of info here and there together, I ended up erasing the beginning and the end of the hard disk. RAID controllers love empty drives, but that is too time consuming to find out if the drive is really bad, or not.
The above commands did reset the Page 0 Damage Error ... finally!
Later, the Vantec Technical Support replied and told me that the error condition is stored into sector 0 of the hard disk, and just erasing the first 10 sectors is sufficient to reset the Page 0 Damage Error condition. I haven't tested this more precise fix myself, but if the problem repeats itself, I'll give it a try.
Dominique.
*****
I added a new storage peripheral to my system a couple of weeks ago, but I just got it to work these last few days. It's a USB3 drive enclosure featuring RAID configurations. It's not directly related to FreeBSD, but I do use the UFS file system with it, and the tips I learned could help a fellow FreeBeastie. It looks like these kind of units are sold under different brands, but many seem to use a JMICRON TECHNOLOGY RAID controller. Mine is the Vantec NexStar HX4R.
While trying to install it, maybe I did something wrong according to Murphey's Law, I ended with a bad drive according to the flashing blue LED of the unit. Moving the hard disk into a different slot made it clear that the problem was with the hard disk, Page 0 Damage Error!
I had doubt about the hard disk being bad because I had it for about 6 months, and never really used it much. How disappointing! A quick read test using the
dd command while the hard disk is plugged into a single drive USB enclosure showed no error whatsoever. dd if=/dev/daX of=/dev/null bs=64k iseek=10 count=1(I use daX throughout this post to prevent disasters in case the ENTER key is pressed before changing the device name. Btw, daX is a 3Tb drive.)
Filled with doubts because a sample test successfully passed doesn't prove the absence of a problem, I continued with the installation of the hard disk manufacturer's full test application. It too didn't find any problem. Me very puzzled, so me do more Internet search digging... Assembling several pieces of info here and there together, I ended up erasing the beginning and the end of the hard disk. RAID controllers love empty drives, but that is too time consuming to find out if the drive is really bad, or not.
dd if=/dev/zero of=/dev/daX bs=64k dd if=/dev/zero of=/dev/daX bs=64k oseek=15261525The above commands did reset the Page 0 Damage Error ... finally!
Later, the Vantec Technical Support replied and told me that the error condition is stored into sector 0 of the hard disk, and just erasing the first 10 sectors is sufficient to reset the Page 0 Damage Error condition. I haven't tested this more precise fix myself, but if the problem repeats itself, I'll give it a try.
Dominique.