February 2, 201511 yr You may want to check the cabling on the new drive and after booting backup, check the smart report on the new drive to see if you see any issues with the drive.
February 3, 201511 yr Author Cabling shouldn't be a problem as I am using a Norco 4224 case with SFF-8087 to SFF-8087 connectors to the backplane. No other drives within that bank are having problems. I started the preclear again last night and it got hung up again on Step 2. This time it only made it 33% of the way through. See the screenshot.
February 3, 201511 yr Community Expert It looks as if the drive has dropped offline for some reason. To get any further diagnosis you really need to provide a syslog that includes the problem period, and ideally a SMART report for the disk in question. The commonest causes of this are always cabling (either power or SATA), or a power supply that cannot handle the number of drives in use. Sometimes it can also be a drive that is not well seated so that vibration can cause a momentary disconnect.
February 3, 201511 yr Author It looks as if the drive has dropped offline for some reason. To get any further diagnosis you really need to provide a syslog that includes the problem period, and ideally a SMART report for the disk in question. The commonest causes of this are always cabling (either power or SATA), or a power supply that cannot handle the number of drives in use. Sometimes it can also be a drive that is not well seated so that vibration can cause a momentary disconnect. I posted the syslog a few posts above the first time that the drive went offline. I have attached the smart reports that I just ran last night for the drive in question. In regard to the power supply, all of the other drives are spun down and not active when the preclear fails. Therefore, I don't believe it to be a power supply issue. smart-short-0202.txt smart-long-0202.txt
February 3, 201511 yr Author Since this preclear problem is off topic from the original post, I have started a new thread found below. Thank you everyone for your help! http://lime-technology.com/forum/index.php?topic=38014.0
February 6, 201511 yr Disk 2 has pending sectors. Disks 4 and 6 have both FAILED end-to-end Error. These three disks cannot be recovered using parity. Recovery software like systemRescueCD can be used to read the disks and recover as much as possible.
February 7, 201511 yr ST3000DM001 - that's he drive BackBlaze had over 50% annual failure rate. I would be replacing those drives ASAP. They're a bit s**t.
February 7, 201511 yr Author Disk 2 has pending sectors. Disks 4 and 6 have both FAILED end-to-end Error. These three disks cannot be recovered using parity. Recovery software like systemRescueCD can be used to read the disks and recover as much as possible. Disk 6 is my failed drive and is being replaced. Why can't these disks be rebuilt using parity?
February 7, 201511 yr Disk 2 has pending sectors. Disks 4 and 6 have both FAILED end-to-end Error. These three disks cannot be recovered using parity. Recovery software like systemRescueCD can be used to read the disks and recover as much as possible. Disk 6 is my failed drive and is being replaced. Why can't these disks be rebuilt using parity? Rebuilding requires all drives except the one being rebuilt to be read flawlessly. If you have more than 1 drive that returns bad or no data at a particular location, the rebuild will fail at that location. See here. http://lime-technology.com/wiki/index.php/FAQ#How_does_parity_work.3F The parity drive is only useful in combination with the other drives.
February 7, 201511 yr Disk 2 has pending sectors. Disks 4 and 6 have both FAILED end-to-end Error. These three disks cannot be recovered using parity. Recovery software like systemRescueCD can be used to read the disks and recover as much as possible. Disk 6 is my failed drive and is being replaced. Why can't these disks be rebuilt using parity? Disk 6 can be rebuilt as long as the other drives do not fail during the procedure. Googling end-to-end error brings up very little concrete information as to what the cause of it is. The safest thing is to just monitor the attribute and replace the drive if it starts increasing. A pending sector will not cause rebuild to fail. The worst case scenario is that a file using that sector *may* be corrupted (and therefore a reconstructed file *may* be corrupted) In this situation, there really isn't another choice but to go ahead
February 7, 201511 yr Disks 4 and 6 have FAILED SMART indicators and should be replaced ASAP. Disk 2 may be useable after a pre-clear.
February 7, 201511 yr Disks 4 and 6 have FAILED SMART indicators and should be replaced ASAP. Disk 2 may be useable after a pre-clear. Absolutely, but you still have to rebuild 6 and hope for the best before you can carry on with replacing the other drives.
Archived
This topic is now archived and is closed to further replies.