April 26, 200818 yr I received two WD 1T drives yesterday, and decided to run them through some tests before inserting into my array. I got one weird reading (see below) If anyone is looking for a decent way to burn in a pair of new disks, you might consider doing something similar. It requires no special tools (except unRAID and smartctl) which you can download here. 1. I put the drives in a test unRAID server with no other disks installed. 2. I captured smartctl output: /boot/smartctl -a -d ata /dev/sda>/boot/smartpre.txt /boot/smartctl -a -d ata /dev/sdb>>/boot/smartpre.txt 3. Through the Web GUI, I made one disk parity and one disk1, formatted disk1, and let it build parity. Worked perfect - no errors in unRAID. (This resulted in a a complete write of parity and complete read of disk1.) 4. Through Web GUI, I ran a parity check - again no errors. (This resulted in a complete read of both disks.) 5. I then switched the disk assignments (parity <-> disk1), and ran steps 3 and 4 again. All ran without a hitch. 6. At this point, each disk has been completely written once, and read 3 complete times. 7. I then reran the smartctl on each drive. /boot/smartctl -a -d ata /dev/sda>/boot/smartpost.txt /boot/smartctl -a -d ata /dev/sdb>>/boot/smartpost.txt There was one anomally - One of the disks had the parameter "raw_read_error_rate" go from 0 to 2. I did some Google searches but couldn't get a good understanding as to what this meant. I found a reference to someone who said if this starts going up by hundreds a day, the drive is about to fail. I saw lots of people with values in the 50s and 60s. I don't think "2" is anything to be overly concerned about, but thought I'd see if anyone here had any thoughts. I don't have many WD drives, but the ones I do have both have this set to 0. Any ideas? Thanks!
April 26, 200818 yr That seems low compared to the drives I've checked here. If my drives have that parameter, they usually have some rather high numbers, although otherwise in good shape. I suspect you'll be watching it over time... It will be interesting to see what trends you find over a month, over a year, etc.
Archived
This topic is now archived and is closed to further replies.