gnollo Posted December 16, 2018 Share Posted December 16, 2018 Running unraid version 6.4.0 I have 1 parity drive + 8 storage drives, running across two hotswapping multidrive dockers. Diagnostics attached. I suspect that there is no issue with power, otherwise it would have affected more than two drives, and with cables (they are all locking cables), will run drive diagnostics too and attach next. tower-diagnostics-20181216-1106.zip Quote Link to comment
gnollo Posted December 16, 2018 Author Share Posted December 16, 2018 smart reports attached tower-smart-20181216-1111.zip tower-smart-20181216-1112.zip Quote Link to comment
JorgeB Posted December 16, 2018 Share Posted December 16, 2018 Both disks dropped offline, reboot so we can see the SMART reports but your server is having issues with several drives, at least 3. Quote Link to comment
trurl Posted December 16, 2018 Share Posted December 16, 2018 35 minutes ago, gnollo said: Diagnostics attached. ... will run drive diagnostics too and attach next. For future reference, Diagnostics already includes SMART for all disks, syslog, and many other useful things. No need to attach them separately. Quote Link to comment
gnollo Posted December 16, 2018 Author Share Posted December 16, 2018 After reboot. Disk 8 and 4 are both in the same 5 drives cage, powered directly by the power supply, no splitting cables used. All the drives are connected via the motherboard own sata ports (8) with the exception of one. Yes, it is drive 8, and yes jonnyblack it was connected to the SAS2LP. I now wish I never bought the thing. What is the best alternative to use when you run out of sata ports, which has better compatibility with unraid? I have swapped 7 and 8 so the SAS2LP is connected to drive 7, in case I rebuild drive 8. Fresh diagnostics attached. tower-diagnostics-20181216-2304.zip Quote Link to comment
WizADSL Posted December 16, 2018 Share Posted December 16, 2018 Assuming you are in the US, here is a good option (this one is already flashed to IT (initiator-target) mode. You will probably need a breakout cable too. https://www.ebay.com/itm/123396360972 Quote Link to comment
gnollo Posted December 16, 2018 Author Share Posted December 16, 2018 Thank you, but should I now rebuild drive 8 or is there a way to bring it back online by assuming that parity is OK and then checking parity against all the drives to confirm said assumption? Quote Link to comment
trurl Posted December 17, 2018 Share Posted December 17, 2018 Are any of your disks actually disabled? Doesn't look like it from the diagnostics, but the diagnostics provided by your very old Unraid version isn't as easy to work with as newer versions. And your plugins are very old too. Please upgrade after you are satisfied that everything is square. Quote Link to comment
gnollo Posted December 17, 2018 Author Share Posted December 17, 2018 Drive 8 is disabled. I am running 6.4.0, interestingly it has not asked me to upgrade to a new version. I had to manually go to the Tools/Update OS and check for update to get notified. Only CA Auto Update Application, Community Applications and Unassigned Devices are looking like they need updating, but this is probably because I am not using the latest version, thanks for pointing that out. How do I rebuild the same drive? Normally I remove the drive and pop another one in to rebuild it. Quote Link to comment
gnollo Posted December 17, 2018 Author Share Posted December 17, 2018 Ok following this tutorial to rebuild the drivehttps://wiki.unraid.net/Replacing_a_Data_Drive Quote Link to comment
gnollo Posted December 17, 2018 Author Share Posted December 17, 2018 I followed the steps twice and upon reboot I find the server back into disk 8 disabled mode. It does not seem to remember I set it to no drive... Quote Link to comment
itimpi Posted December 17, 2018 Share Posted December 17, 2018 15 minutes ago, gnollo said: I followed the steps twice and upon reboot I find the server back into disk 8 disabled mode. It does not seem to remember I set it to no drive... If the drive is not disabled then you first have to start the array with it set to no drive. When you then stop the array and assign a new drive it will start the rebuild. Quote Link to comment
gnollo Posted December 17, 2018 Author Share Posted December 17, 2018 If the drive is not disabled then you first have to start the array with it set to no drive. When you then stop the array and assign a new drive it will start the rebuild.Done that twice, keeps automatically reassigning it. I'll pull the drive physically out next time, and use a different drive to rebuild (got a new spare drive)Sent from my SM-A520F using Tapatalk Quote Link to comment
itimpi Posted December 17, 2018 Share Posted December 17, 2018 3 minutes ago, gnollo said: Done that twice, keeps automatically reassigning it. I'll pull the drive physically out next time, and use a different drive to rebuild (got a new spare drive) Sent from my SM-A520F using Tapatalk That is strange! Are you sure you got the sequence right? Stop the array (if it is started) Set the drive to not be assigned Start the array (it should warn you about it being unprotected). This step makes it forget the current assignment. Stop the array Assign the drive you want to rebuild onto Start the array (it should now be warning you that a rebuild will take place). The only thing I can think that might stop that sequence working would be if there are problems with the flash drive so that Unraid cannot save the new settings. Quote Link to comment
JorgeB Posted December 17, 2018 Share Posted December 17, 2018 Didn't have the chance to reply earlier, but like mentioned server is showing issues with several disks: disk8 - unfortunately SMART is disabled for the disk, so no SMART report when outside the array, you need to enable SMART: smartctl -s on /dev/sdg then grab and post a report. disk4 - dropped offline and is on the onboard SATA, so not SAS2LP related, it does show a pending sector but the way it dropped seems unrelated and it's possibly a false positive, still good to run an extended SMART test to confirm and replace cables to rule them out if the same happens again. disk2 - there was a timeout error, replace cables. P.S. the onboard Intel controller is set to IDE, change it to AHCI. Quote Link to comment
gnollo Posted December 17, 2018 Author Share Posted December 17, 2018 1 hour ago, itimpi said: That is strange! Are you sure you got the sequence right? Stop the array (if it is started) Set the drive to not be assigned Start the array (it should warn you about it being unprotected). This step makes it forget the current assignment. Stop the array Assign the drive you want to rebuild onto Start the array (it should now be warning you that a rebuild will take place). The only thing I can think that might stop that sequence working would be if there are problems with the flash drive so that Unraid cannot save the new settings. I was following the process above - Stop the array - Unassign the old drive if still assigned (to unassign, set it to No Device) - power down I will try yours next, after I have addressed the points raised by johnnie.black Quote Link to comment
itimpi Posted December 17, 2018 Share Posted December 17, 2018 Just now, gnollo said: I was following the process above - Stop the array - Unassign the old drive if still assigned (to unassign, set it to No Device) - power down I will try yours next, after I have addressed the points raised by johnnie.black That will not work - you have to have started the array to get it to remember the change as it is only on array start that disk assignments are saved. Quote Link to comment
gnollo Posted December 17, 2018 Author Share Posted December 17, 2018 1 hour ago, johnnie.black said: Didn't have the chance to reply earlier, but like mentioned server is showing issues with several disks: disk8 - unfortunately SMART is disabled for the disk, so no SMART report when outside the array, you need to enable SMART: smartctl -s on /dev/sdg then grab and post a report. disk4 - dropped offline and is on the onboard SATA, so not SAS2LP related, it does show a pending sector but the way it dropped seems unrelated and it's possibly a false positive, still good to run an extended SMART test to confirm and replace cables to rule them out if the same happens again. disk2 - there was a timeout error, replace cables. P.S. the onboard Intel controller is set to IDE, change it to AHCI. I executed the command, new diagnostics attached. Disk 4 has had a pending sector for a very long time, I have another drive ready to replace it after I fix this. Disk 2: are there any brand sata cable you would recommend? Intel controller: will change it to AHCI this evening, when I am back, as I guess i have to do that in the BIOS tower-diagnostics-20181217-1020.zip Quote Link to comment
gnollo Posted December 17, 2018 Author Share Posted December 17, 2018 9 minutes ago, itimpi said: That will not work - you have to have started the array to get it to remember the change as it is only on array start that disk assignments are saved. Thanks! Then what is indicated in this tutorial is wrong and needs to be updated https://wiki.unraid.net/Replacing_a_Data_Drive Quote Link to comment
JorgeB Posted December 17, 2018 Share Posted December 17, 2018 Disk8 looks fine, could be cables, could be the SAS2LP Disk4 should also be fine for now, problem looks connection related. Don't use any specific cable brand, just replace them so you can rule them out, also good to replace or at least swap the power cables on those disks. 45 minutes ago, gnollo said: Thanks! Then what is indicated in this tutorial is wrong and needs to be updated You're not replacing a disk, you're re-enabling it, there's also a wiki about that: https://wiki.unraid.net/Troubleshooting#Re-enable_the_drive Quote Link to comment
gnollo Posted December 17, 2018 Author Share Posted December 17, 2018 2 hours ago, johnnie.black said: Disk8 looks fine, could be cables, could be the SAS2LP Disk4 should also be fine for now, problem looks connection related. Don't use any specific cable brand, just replace them so you can rule them out, also good to replace or at least swap the power cables on those disks. You're not replacing a disk, you're re-enabling it, there's also a wiki about that: https://wiki.unraid.net/Troubleshooting#Re-enable_the_drive RE: power cables, I use a Norco SS500, which houses another 2 drives, which showed no problems. Does that not eliminate a power issue? Quote Link to comment
JorgeB Posted December 17, 2018 Share Posted December 17, 2018 12 minutes ago, gnollo said: Does that not eliminate a power issue? Most likely, though it could be an issue with the backplane, replace the SATA cables and you can also swap the disks around so in case it happens again you can see if the issues followed the disk or stayed with the slot. Quote Link to comment
gnollo Posted December 18, 2018 Author Share Posted December 18, 2018 Well, I changed the setting to AHCI, said YES to saying BIOS, screen went blank and the server never came back up. Lights come on, fans are spinning but I get no picture whatsoever on the screen upon reboot. Any suggestions on how to reset the motherboard to previous settings? Quote Link to comment
gnollo Posted December 18, 2018 Author Share Posted December 18, 2018 Motherboard is a Gigabyte Technology G33-DS3R/G33-DS3R, BIOS F7L 07/31/2009 Quote Link to comment
gnollo Posted December 18, 2018 Author Share Posted December 18, 2018 I guess if I pop out the battery with the power cable removed, and wait a couple of hours, that should do it. What other settings are important for me to save in the BIOS to get the best out of Unraid? I will try to set AHCI again but this time it will be the first setting I edit and save. Quote Link to comment
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.