September 30, 201015 yr I recently purchased 2 of the Hitachi 2TB drives after getting an email alert from Newegg of a $99/drive deal. These drives also seemed to get a favorable response from the unRAID community based on comments in the 'Deals' forum. I currently have 5 WD 1TB EAC drives (1 parity + 4 data), which were working great. I wanted to expand the system, so I needed to replace the parity and disk1 drives with the new 2TB drives. I also plan to add one of the left over data drives to the array, which will bring my total of data drives to 5. Since I only have the unRAID Server Plus license, that is the max number of drives I can have. Yesterday, I swapped out the 1TB WD parity drive and installed the new 2TB Hitachi drive. It took about 9 hours to complete the parity sync. Everything went great, there were no errors. This morning, I swapped out the 1TB WD disk1 drive and installed the 2nd 2TB Hitachi drive. I am currently at about 34.2% of the data-rebuild process. I now noticed that the parity drive is showing 151 errors. So far, no errors are listed for the new 2TB disk1. Is it normal to get parity drive errors during this data-rebuild process? Is this a reasonable amount of errors (I've never had any errors with the WD drives)? Or, could I have a defective drive? If so, how do I go about removing it and swapping it with a replacement? Will I need to put back the 1TB parity and the 1TB disk1 drives? Then send the drive back for a replacement?
September 30, 201015 yr I recently purchased 2 of the Hitachi 2TB drives after getting an email alert from Newegg of a $99/drive deal. These drives also seemed to get a favorable response from the unRAID community based on comments in the 'Deals' forum. I currently have 5 WD 1TB EAC drives (1 parity + 4 data), which were working great. I wanted to expand the system, so I needed to replace the parity and disk1 drives with the new 2TB drives. I also plan to add one of the left over data drives to the array, which will bring my total of data drives to 5. Since I only have the unRAID Server Plus license, that is the max number of drives I can have. Yesterday, I swapped out the 1TB WD parity drive and installed the new 2TB Hitachi drive. It took about 9 hours to complete the parity sync. Everything went great, there were no errors. This morning, I swapped out the 1TB WD disk1 drive and installed the 2nd 2TB Hitachi drive. I am currently at about 34.2% of the data-rebuild process. I now noticed that the parity drive is showing 151 errors. So far, no errors are listed for the new 2TB disk1. Is it normal to get parity drive errors during this data-rebuild process? Is this a reasonable amount of errors (I've never had any errors with the WD drives)? Or, could I have a defective drive? If so, how do I go about removing it and swapping it with a replacement? Will I need to put back the 1TB parity and the 1TB disk1 drives? Then send the drive back for a replacement? Errors are not normal, they indicate a "read" error. Normally, this is not an issue, but in a re-construction of a data drive, they are. Did you perform a parity check after you initially calculated parity on the new drive? If not, it was never read, only written. I suspect you did not perform that step. If you had, then those un-readable sectors would probably been re-allocated and parity correctly stored. It can't do that when re-constructing a drive in a replacement process. I'm guessing you did not pre-clear the new drives either... that would have tested their ability to read and write to all the sectors, again possibly showing the bad sectors before you trusted them to your data. So... I would do this: Let the current re-construction complete. It probably has some bad data where parity could not be read. Post a copy of your syslog. (Instructions under troubleshooting in the wiki) It might have some clues. Get a copy of a smart report for the new parity drive. smartctl -d ata -a /dev/sdX where sdX = the device of your parity drive. Post the results. Then, stop the server, power down, put the original 1TB data disk back into place, possibly put the original parity disk back too. power up. The array will not start, as it will see the disk changes. Log in via telnet and type initconfig Respond with "Yes" to its prompt (Capital "Y" lower case "es") Refresh your management console. All the indicators should be blue. Start your array. It will calculate parity as if it is doing it the first time. When it is done, you'll be back the way you were before you started swapping around disks. Before you do ANYTHING ELSE, do a full parity check. This will attempt to read the parity you just wrote. It is, I think, the step you did not perform the first time. (You did not mention it) When it is done it will be as if you never tried upgrading to the 2TB disks. Then, if you have a spare disk controller port, stop the array, power down, connect the new "parity" disk to the spare port. Do not assign it to the array. (In the same way, if you have two spare ports attach both new drives... but do not assign them) Power up... The array will be as it originally was, with your older disks in place. If you do not have a spare port let us know, we'll suggest alternatives. If you do, check out the preclear_disk.sh script. Specifically designed to burn-in disks and discover issues before you assign the disks in the array. Basically.. next time. Step 1. Do a full parity check BEFORE anything else, to make sure no other issues exist, before swapping in any new disk. Step 2. Run the preclear_disk.sh script on each of the new disks (if at all possible, to burn them in, even if you do this on a different PC) Step 3. Power down, replace the parity disk, power up, re-construct parity onto it. Step 4. Do a full parity check... Step 3 wrote parity, but you'll not know if it can be read until a full parity check is performed. Step 5. (optional, but if not pressed for time) Do another parity check, or two, or three. Step 6. Stop the array, power down, replace the data disk. Power up, let it rebuild by pressing the "Start" button. Step 7. Again we just wrote the data disk, we'll not know for sure it can be read back until you do one more full parity check. Joe L.
September 30, 201015 yr Author Did you perform a parity check after you initially calculated parity on the new drive? No, I followed the instructions in the unRAID Manual: Replace a single disk with a bigger one This is the case where you are replacing a single small disk with a bigger one: 1. Stop the array. 2. Power down the unit. 3. Replace smaller disk with new bigger disk. 4. Power up the unit. 5. Start the array. If you are replacing your existing Parity disk with a bigger one, then when you Start the array, the system will simply start a parity sync onto the new Parity disk. I did not see any instruction to run a parity check after the parity sync. I guess I assumed the parity sync would also perform a parity check. After following those instruction, my console showed everything was fine and no errors. Then, I repeated the process for the disk1 drive, which is still on-going and probably won't be complete for another 6-1/2 hours. In the case of replacing a data disk with a larger one, the manual says: When you start the array, the system will reconstruct the contents of the original smaller disk onto the new disk. Upon completion, the disk's file system will be expanded to reflect the new size. You can only expand one disk at a time. Again, it doesn't say anything about running a parity check. I'm guessing you did not pre-clear the new drives either... that would have tested their ability to read and write to all the sectors, again possibly showing the bad sectors before you trusted them to your data. And no, I did not pre-clear the drives. Let the current re-construction complete. It probably has some bad data where parity could not be read. Post a copy of your syslog. I've attached a copy of my syslog during the re-construction and will post another after re-construction is complete. Get a copy of a smart report for the new parity drive. smartctl -d ata -a /dev/sdX where sdX = the device of your parity drive. Post the results. I assume I need to login via Telnet and type in this command, right? What do you mean by "sdx = the device of your parity drive"? Is that the long string listed on the device tab of the console? In my case it is 'pci-0000:00:1f.2-scsi-0:0:0:0 host3 (sda) Hitachi_HDS722020ALA330_JK1151YAJ69VEZ' Is that what you are referring to? After that, I'll follow the rest of your instructions and will post any additional questions. I do have a couple spare SATA ports, so I should be able to run the preclear_disk.sh script. Darn, I thought this was going to be a lot easier. I was hoping to have my unRAID server fully functional by tomorrow, so I could rip a couple new DVDs to it. Oh well, I can still play the DVDs from the DVD drive and rip later. Maybe we should edit the manual to include these important steps. Thanks a lot for your help. syslog-2010-09-30-1537.txt
September 30, 201015 yr Author I assume I need to login via Telnet and type in this command, right? What do you mean by "sdx = the device of your parity drive"? Is that the long string listed on the device tab of the console? In my case it is 'pci-0000:00:1f.2-scsi-0:0:0:0 host3 (sda) Hitachi_HDS722020ALA330_JK1151YAJ69VEZ' Nevermind, I figured it out. Attached is a copy of my smart report for the parity drive. smart.txt
September 30, 201015 yr Did you perform a parity check after you initially calculated parity on the new drive? No, I followed the instructions in the unRAID Manual: Replace a single disk with a bigger one This is the case where you are replacing a single small disk with a bigger one: 1. Stop the array. 2. Power down the unit. 3. Replace smaller disk with new bigger disk. 4. Power up the unit. 5. Start the array. If you are replacing your existing Parity disk with a bigger one, then when you Start the array, the system will simply start a parity sync onto the new Parity disk. I did not see any instruction to run a parity check after the parity sync. I guess I assumed the parity sync would also perform a parity check. After following those instruction, my console showed everything was fine and no errors. Then, I repeated the process for the disk1 drive, which is still on-going and probably won't be complete for another 6-1/2 hours. Unfortunately, you assumed incorrectly, and even though the parity "check" afterward is not needed if everything goes right, and the disks are perfect, it is recommended by the user-community based on our collective experiences. In the case of replacing a data disk with a larger one, the manual says: When you start the array, the system will reconstruct the contents of the original smaller disk onto the new disk. Upon completion, the disk's file system will be expanded to reflect the new size. You can only expand one disk at a time. Again, it doesn't say anything about running a parity check. Again, this is based on the collective experiences of us users. The reading of the new drive is not tested until it first occurs. It does not occur when it is re-constructed from the other disks, therefore, the first time you'll be reading your brand new drive is on the first parity check. I'm guessing you did not pre-clear the new drives either... that would have tested their ability to read and write to all the sectors, again possibly showing the bad sectors before you trusted them to your data. And no, I did not pre-clear the drives. Let the current re-construction complete. It probably has some bad data where parity could not be read. Post a copy of your syslog. I've attached a copy of my syslog during the re-construction and will post another after re-construction is complete. Get a copy of a smart report for the new parity drive. smartctl -d ata -a /dev/sdX where sdX = the device of your parity drive. Post the results. I assume I need to login via Telnet and type in this command, right? What do you mean by "sdx = the device of your parity drive"? Is that the long string listed on the device tab of the console? In my case it is 'pci-0000:00:1f.2-scsi-0:0:0:0 host3 (sda) Hitachi_HDS722020ALA330_JK1151YAJ69VEZ' Is that what you are referring to? The device is the three letter designation in the parens. In this case "sda" So... the command would be smartctl -d ata -a /dev/sda Yes, you need to log in via telnet to type that command. You do not need to wait for the re-construction to finish as it will not have any effect on it. After that, I'll follow the rest of your instructions and will post any additional questions. I do have a couple spare SATA ports, so I should be able to run the preclear_disk.sh script. Darn, I thought this was going to be a lot easier. I was hoping to have my unRAID server fully functional by tomorrow, so I could rip a couple new DVDs to it. Oh well, I can still play the DVDs from the DVD drive and rip later. True... Maybe we should edit the manual to include these important steps. Unfortunately, we do not have the authority to edit the official manual, only the un-official user-edited section. and based on your feedback, I just added those notes to the un-official copy of the manual. http://lime-technology.com/wiki/index.php?title=Un-Official_UnRAID_Manual#Replace_a_single_disk_with_a_bigger_one Thanks a lot for your help. You are welcome Joe L.
September 30, 201015 yr I assume I need to login via Telnet and type in this command, right? What do you mean by "sdx = the device of your parity drive"? Is that the long string listed on the device tab of the console? In my case it is 'pci-0000:00:1f.2-scsi-0:0:0:0 host3 (sda) Hitachi_HDS722020ALA330_JK1151YAJ69VEZ' Nevermind, I figured it out. Attached is a copy of my smart report for the parity drive. It shows one sector has already been re-allocated, and three others are pending re-allocation when next written. Overall, this is not bad, as disks have several thousand spare sectors, but they are bad when trying to re-construct onto a new disk as you were doing. The preclear_disk.sh script is specifically designed to test the disk for these types of problems before you put the disks in your array. It is also the only way to avoid an extended down-time once you do have a parity drive assigned. Otherwise, the array would be of-line for 6 or more hours while a 2TB drive is fully written with zeros. I highlighted the errors in red. The "raw" value is what we are looking at for those two attributes. For almost all the other, we only look at the normalized "VALUE" and compare it to the failure "THRESH" to know the status. If the VALUE goes below the THRESH, the drive is failing a SMART test. ID# ATTRIBUTE_NAME FLAG VALUE WORST THRESH TYPE UPDATED WHEN_FAILED RAW_VALUE 1 Raw_Read_Error_Rate 0x000b 100 100 016 Pre-fail Always - 0 2 Throughput_Performance 0x0005 133 133 054 Pre-fail Offline - 101 3 Spin_Up_Time 0x0007 100 100 024 Pre-fail Always - 484 4 Start_Stop_Count 0x0012 100 100 000 Old_age Always - 6 5 Reallocated_Sector_Ct 0x0033 100 100 005 Pre-fail Always - 1 7 Seek_Error_Rate 0x000b 100 100 067 Pre-fail Always - 0 8 Seek_Time_Performance 0x0005 119 119 020 Pre-fail Offline - 36 9 Power_On_Hours 0x0012 100 100 000 Old_age Always - 24 10 Spin_Retry_Count 0x0013 100 100 060 Pre-fail Always - 0 12 Power_Cycle_Count 0x0032 100 100 000 Old_age Always - 5 192 Power-Off_Retract_Count 0x0032 100 100 000 Old_age Always - 6 193 Load_Cycle_Count 0x0012 100 100 000 Old_age Always - 6 194 Temperature_Celsius 0x0002 200 200 000 Old_age Always - 30 (Lifetime Min/Max 23/33) 196 Reallocated_Event_Count 0x0032 100 100 000 Old_age Always - 1 197 Current_Pending_Sector 0x0022 100 100 000 Old_age Always - 3 198 Offline_Uncorrectable 0x0008 100 100 000 Old_age Offline - 0 199 UDMA_CRC_Error_Count 0x000a 200 200 000 Old_age Always - 0
September 30, 201015 yr Author So, I'm now planning to do the following: 1. Let the current data-rebuild process complete. 2. Generate a new copy of my syslog and post to this thread. 3. Stop the array / power down the server. 4. Remove the 2TB parity and 2TB disk1 drives 5. Re-install the original 1TB parity and 1TB disk1 drives 6. Power-up the server 7. Log in via telnet and type 'initconfig', respond with "Yes" to its prompt 8. Using the browser on another machine on the network, refresh the unRAID management console (all indicators should be blue) 9. Start the array 10. After it finishes calculating parity, do a full parity check 11. Stop the array / power down the server. 12. Connect both of the 2TB drives to the spare SATA ports on the server's mobo. 12. Power-up the server 13. Log in via telnet 14. type 'cd /boot' 15. type 'preclear_dsk.sh /dev/??? (where is the 2TB parity device, such as hda, hdb, sda, sdb, etc.) 16. repeat step 15 for the 2TB data (disk1) drive 17. Stop the array / power down the server. 18. Swap out the 1TB parity drive for the 2TB parity drive. 19. Power-up the server 20. Reconstruct parity 21. Do full parity check 22. Optional - repeat step 21 a couple more times 23. Stop the array / power down the server. 24. Swap out the 1TB data (disk1) drive for the 2TB data drive. 25. Rebuild the data disk1 by pressing the 'Start' button 26. Do full parity check Question for step #15: I won't know what to type in for the preclear command until I see what the IDs are for the 2TB drives, will this be displayed in the 'devices' page of the unRAID console? Question for step #22: How critical is this, and why 2 or 3 extra times? How long will this take? Thanks again Joe for your help!! -Gregg
September 30, 201015 yr Question for step #15: I won't know what to type in for the preclear command until I see what the IDs are for the 2TB drives, will this be displayed in the 'devices' page of the unRAID console? Yes, they will be in parens in the drop-down list. or you could type ls -l /dev/disk/by-id The disks will be listed there by model/serial number with the device name at the end of the line. Disks with partitions will also have entries ending with a "1" You want the three letter designation. (sda, sdb, hda, hdb, etc) Question for step #22: How critical is this, and why 2 or 3 extra times? How long will this take? Not critical, but it just lets you get more hours on the drive before putting your data on it. To weed out early mechanical and electrical failures) Thanks again Joe for your help!! -Gregg obviously, you need to download the preclear_disk.sh and un-zip it to your flash share, otherwise it would not be there when you go to invoke it. Also, you can start multiple telnet sessions, and a pre-clear in each. The pre-clear of a 2TB drive takes roughly 30 to 35 hours. It reads the entire disk, then writes it, the re-reads it comparing what it is reading for all zeros. It also reads many random sectors at the same time it ir reading the disk linearly, attempting to move the disk heads far more than usual when just playing a movie. Again, this is to get the disk to show mechanical issues before trusting your data to it. Just writing zeros regardless of how takes nearly 6 hours or more. You'll be able to track its progress on the screen. You'll need to leave the telnet session open and active for the entire time, so if that is not possible, either do it on the system console or install "screen" and use it to manage your screen sessions so you can log off without the process terminating. Joe L.
September 30, 201015 yr So, I'm now planning to do the following: 1. Let the current data-rebuild process complete. 2. Generate a new copy of my syslog and post to this thread. 3. Stop the array / power down the server. 4. Remove the 2TB parity and 2TB disk1 drives 5. Re-install the original 1TB parity and 1TB disk1 drives 6. Power-up the server 7. Log in via telnet and type 'initconfig', respond with "Yes" to its prompt 8. Using the browser on another machine on the network, refresh the unRAID management console (all indicators should be blue) 9. Start the array 10. After it finishes calculating parity, do a full parity check 11. Stop the array / power down the server. 12. Connect both of the 2TB drives to the spare SATA ports on the server's mobo. 12. Power-up the server 13. Log in via telnet 14. type 'cd /boot' 15. type 'preclear_dsk.sh /dev/??? (where is the 2TB parity device, such as hda, hdb, sda, sdb, etc.) 16. repeat step 15 for the 2TB data (disk1) drive 17. Stop the array / power down the server. 18. Swap out the 1TB parity drive for the 2TB parity drive. 19. Power-up the server 20. Reconstruct parity 21. Do full parity check 22. Optional - repeat step 21 a couple more times 23. Stop the array / power down the server. 24. Swap out the 1TB data (disk1) drive for the 2TB data drive. 25. Rebuild the data disk1 by pressing the 'Start' button 26. Do full parity check Question for step #15: I won't know what to type in for the preclear command until I see what the IDs are for the 2TB drives, will this be displayed in the 'devices' page of the unRAID console? Question for step #22: How critical is this, and why 2 or 3 extra times? How long will this take? Thanks again Joe for your help!! -Gregg if you can take about 36 hours window without parity protection, here is what you can do to speed up. (a) un-assign 2 TB parity disk. Your unRAID now has no protection. (b) preclear this 2TB un-assigned parity disk. (this will take about 24 hours) © Once preclear is done, shutdown machine and put 1TB data disk back to original assigned slot. (d) Power on machine, assign 2TB parity disk back to parity slot. Invalid current configuration to create a new configuration and kick off parity re-genration with 2TB parity disk. (e) Once parity regenration is done (this process should take no more than 12 hours), you have protection. (f) now preclear new 2TB data disk. (g) Once precleear is done, stop unRAID and un-assign 1TB data disk you are going to replace with this clean 2TB data disk assigned to same slot. (h) Kick off data regeneration on this newly assigned 2TB data disk. (i) once data regeneration is finished. shut down machine, take out replaced 1TB data disk. (j) Power on machine and browsing through data to make sure everything is fine.
October 1, 201015 yr Author if you can take about 36 hours window without parity protection, here is what you can do to speed up. I don't think I can go 36 hours without protection, so I'll follow Joe's instructions. Thanks anyhow.
October 1, 201015 yr Author It finally completed the data-rebuild process and the lights are all green. The process completed with only the 151 errors that I originally reported. It must have happened sometime between 24% and 34% complete. Now I'm going to follow through with the rest of the instructions. Step 2 - generate new syslog. Attached. syslog-2010-09-30-2129.txt
October 1, 201015 yr Author You'll be able to track its progress on the screen. You'll need to leave the telnet session open and active for the entire time, so if that is not possible, either do it on the system console or install "screen" and use it to manage your screen sessions so you can log off without the process terminating. I was under the impression that Putty would shut down the telnet session automatically if it times out. Will this happen during the preclear routine? When you said "do it on the system console," are referring to the server's display screen? Since this is a headless server, I don't have a monitor to look at, but I suppose I could pull one from another system. What is "screen"? I'd be happy to install it, if you can point me to a link. Thanks again, Gregg
October 1, 201015 yr screen (screen manager with VT100/ANSI terminal emulation) If you have unMENU installed, you can go to the Pkg Manager screen and scroll down to find it. Then you are able to download/install it from there.
October 1, 201015 yr You'll be able to track its progress on the screen. You'll need to leave the telnet session open and active for the entire time, so if that is not possible, either do it on the system console or install "screen" and use it to manage your screen sessions so you can log off without the process terminating. I was under the impression that Putty would shut down the telnet session automatically if it times out. Will this happen during the preclear routine? No, it will stay active as long as the LAN stays running, and the server keeps running (a UPS helps). You might also disable any automatic screensaver/go to sleep on your PC in case it decides you've not typed anything in a while and decides to turn itself off. When you said "do it on the system console," are referring to the server's display screen? Since this is a headless server, I don't have a monitor to look at, but I suppose I could pull one from another system. I understand. For some tasks, the systems console is easiest. There are actual 6 virtual system consoles, normally you only use the first one, but you can switch between them with ALT-F1 through ALT-F6. Each will let you log on and run a separate pre-clear process. What is "screen"? I'd be happy to install it, if you can point me to a link. Here is a link: http://lime-technology.com/forum/index.php?topic=2817.msg24827#msg24827 Thanks again, Gregg You are welcome. You'll need to re-install screen each time you reboot. (it will already be downloaded, so you'll just need to cd /boot/packages and then type the two installpkg commands listed above.
October 2, 201015 yr Author The pre-clear completed for both drives earlier this morning. Both of them had this displayed on the PuTTY screen near the end.: =========================================================================== = unRAID server Pre-Clear disk /dev/sda = cycle 1 of 1 = Disk Pre-Clear-Read completed DONE = Step 1 of 10 - Copying zeros to first 2048k bytes DONE = Step 2 of 10 - Copying zeros to remainder of disk to clear it DONE = Step 3 of 10 - Disk is now cleared from MBR onward. DONE = Step 4 of 10 - Clearing MBR bytes for partition 2,3 & 4 DONE = Step 5 of 10 - Clearing MBR code area DONE = Step 6 of 10 - Setting MBR signature bytes DONE = Step 7 of 10 - Setting partition 1 to precleared state DONE = Step 8 of 10 - Notifying kernel we changed the partitioning DONE = Step 9 of 10 - Creating the /dev/disk/by* entries DONE = Step 10 of 10 - Testing if the clear has been successful. DONE = Disk Post-Clear-Read completed DONE Disk Temperature: 30C, Elapsed Time: 28:10:44 ============================================================================ == == Disk /dev/sda has been successfully precleared == ============================================================================ I'm not sure how to interpret the results, so I copied the remain text to a txt file. I have 2 reports, one for the "sdg" drive and the other for the "sda" drive. It looks like the "sdg" drive has some problems and the "sda" drive is okay. Could someone please take a look and confirm this? I've attached these files to this post. I'm confused because it looks like the script for the "sdg" drive ended with a S.M.A.R.T. error test and the scipt for the "sdg" drive just tried repeating another cycle of the complete test, but didn't work based on seeing a lot of "-bash: =: command not found" strings of text. post_Pre-Clear_sdg.txt post_Pre-Clear_sda.txt
October 2, 201015 yr Author The reports in the previous post are cut&pastes from the PuTTY session screens. Later, I found that you can copy everything from a PuTTY session to the clipboard with a menu command, so to be more complete, I went ahead and did that. I've attached these to this message if more info is needed. BTW, Does anyone know why the Pre-Clear session for the sda drive shows 8 99% progress completes before it finally reported the pre-clears successfully completed? The sdg just shows that it successfully completed the preclear. Thanks in advance for your help. -Gregg Pre-Clear_session_for_sdg_drive.txt Pre-Clear_session_for_sda_drive.txt
October 3, 201015 yr I looks like you attempted to paste the contents of the output back into the telnet window. Since each line started with a "= sign, and there is no command named "=" or "==" none were found. (in telnet, when you highlight text, it is automatically copied to the clipboard, and a right-mouse click then pastes it.) Attach a copy of your syslog to the next post. It will have the full output of the pre-clear processes.
October 3, 201015 yr Author Damn, yeah I tried using normal Windows copy & paste techniques out of habit, I remember originally highlighting and right clicking over an area. Then I noticed that the cursor jumped to the end. So I just used Ctrl-C after that. I didn't realize that my attempts to copy were pasting back into the telnet window. Unfortunately, we had a power failure (a very rare occurrence) at our house last night, so everything shut down, including the PC from which I was running PuTTY and my unRAID server. I'm glad the power failure didn't occur during the preclear, but not knowing what caused it I powered up the server and it booted up ok, but now unRAID is going through a Parity-Check. Isn't the syslog cleared if you re-start? Is there any way to decipher the mess, or should I try running preclear again?
October 3, 201015 yr I powered up the server and it booted up ok, but now unRAID is going through a Parity-Check. That is quite normalIsn't the syslog cleared if you re-start?Yes, it is in memory, so when you lost power the old syslog was lost. Is there any way to decipher the mess, or should I try running preclear again? You can test if it is pre-cleared by typing preclear_disk.sh -t /dev/sdX where sdX = the device under test. Consider getting a UPS. (APC brand is easiest to configure for most unRAID users)
October 3, 201015 yr Author Both disks are pre-cleared. Media login: root Linux 2.6.32.9-unRAID. root@Media:~# cd /boot root@Media:/boot# preclear_disk.sh -t /dev/sdg Pre-Clear unRAID Disk ######################################################################## Device Model: Hitachi HDS722020ALA330 Serial Number: JK1151YAJ69VEZ Firmware Version: JKAOA3EA User Capacity: 2,000,398,934,016 bytes Disk /dev/sdg: 2000.3 GB, 2000398934016 bytes 255 heads, 63 sectors/track, 243201 cylinders, total 3907029168 sectors Units = sectors of 1 * 512 = 512 bytes Disk identifier: 0x00000000 Device Boot Start End Blocks Id System /dev/sdg1 63 3907029167 1953514552+ 0 Empty Partition 1 does not end on cylinder boundary. ######################################################################## ============================================================================ == == DISK /dev/sdg IS PRECLEARED == ============================================================================ root@Media:/boot# preclear_disk.sh -t /dev/sda Pre-Clear unRAID Disk ######################################################################## Device Model: Hitachi HDS722020ALA330 Serial Number: JK1151YAJ69URZ Firmware Version: JKAOA3EA User Capacity: 2,000,398,934,016 bytes Disk /dev/sda: 2000.3 GB, 2000398934016 bytes 255 heads, 63 sectors/track, 243201 cylinders, total 3907029168 sectors Units = sectors of 1 * 512 = 512 bytes Disk identifier: 0x00000000 Device Boot Start End Blocks Id System /dev/sda1 63 3907029167 1953514552+ 0 Empty Partition 1 does not end on cylinder boundary. ######################################################################## ============================================================================ == == DISK /dev/sda IS PRECLEARED == ============================================================================ root@Media:/boot# What about the errors displayed right after the pre-cleard finished with the "sdg" drive? There were several, such as: < ATA Error Count: 20 (device log contains only the most recent five errors) --- > ATA Error Count: 63 (device log contains only the most recent five errors) 84c84 < Error 20 occurred at disk power-on lifetime: 22 hours (0 days + 22 hours) --- > Error 63 occurred at disk power-on lifetime: 31 hours (1 days + 7 hours) 90c90 and 101c101 < Error 19 occurred at disk power-on lifetime: 22 hours (0 days + 22 hours) --- > Error 62 occurred at disk power-on lifetime: 31 hours (1 days + 7 hours) 107c107 < 40 51 ae 01 b8 ee 0c Error: UNC 174 sectors at LBA = 0x0ceeb801 = 21697 1265 118c118 < Error 18 occurred at disk power-on lifetime: 22 hours (0 days + 22 hours) --- > Error 61 occurred at disk power-on lifetime: 31 hours (1 days + 7 hours) 124c124 < 40 51 ae 01 b8 ee 0c Error: UNC 174 sectors at LBA = 0x0ceeb801 = 21697 1265 etc., etc...
October 3, 201015 yr You yourself included the line that describes what they represent. They are just more detail describing (to the manufacturer) the un-readable sectors 40 51 ae 01 b8 ee 0c Error: UNC 174 sectors at LBA = 0x0ceeb801 = 21697
October 6, 201015 yr Author After finishing pre-clearing both my new Hitachi 2TB drives, I've installed one into the parity slot and it is currently reconstructing parity. I will follow this with several parity checks. Next, I plan to swap out the 1TB disk1 data drive for the other new 2TB Hitachi drive. After rebuilding the data, I will do another full parity check. I now have 2 - 1TB WD spare drives. I plan to put one of these in the disk5 position to add to my array. Both of the drives I'm removing have been running flawlessly for over a year. Would it help to run pre-clear on the disk I eventual add to the array? Should I use the -n option to skip pre-read and post-read?
October 6, 201015 yr You should have left the drives in the array to avoid unnecessary nuisance of adding it to the array later. I would recommend preclearing it to avoid lengthy array downtime. Pre and post reads likely unnecessary but wouldn't hurt. I'd likely let them run.
Archived
This topic is now archived and is closed to further replies.