September 22, 201015 yr Today, for no apparent reason I lost connection to my unRaid box. I was not able to see it on a network and could not telnet in to it. Monitor connected to it was also blank. The machine is on a UPS and there have been no power outages. I saw no other alternative, so I hard booted the computer down and then back up. During the boot I saw a lot of "unable to enumerate USB device on port" (or something similar to that) messages on the screen and now parity check is running. Any thoughts? Thanks, Lev unRAID 4.5.4 pro 8 data drives, + cache. syslog-2010-09-21_a.txt.zip
September 22, 201015 yr You might need to re-upload your syslog. The formatting in it is all wrong, which makes it very hard to read. What program did you use to capture it? Notepad should be fine (assuming you are using Windows). Also, you shouldn't need to zip it. If it is too large to attach to a post here, put it on pastebin.
September 22, 201015 yr Looks like a USB hub, or something connected to it is failing. Sep 21 22:58:56 Tower kernel: hub 2-0:1.0: unable to enumerate USB device on port 1 Sep 21 22:59:27 Tower last message repeated 122 times Sep 21 23:00:28 Tower last message repeated 242 times Sep 21 23:01:29 Tower last message repeated 241 times Sep 21 23:02:08 Tower last message repeated 156 times Sep 21 23:02:08 Tower in.telnetd[2326]: connect from 192.168.1.133 (192.168.1.133) Sep 21 23:02:08 Tower kernel: hub 2-0:1.0: unable to enumerate USB device on port 1 Sep 21 23:02:16 Tower last message repeated 30 times Sep 21 23:02:16 Tower login[2327]: invalid password for `UNKNOWN' on `pts/0' from `192.168.1.133' Sep 21 23:02:16 Tower kernel: hub 2-0:1.0: unable to enumerate USB device on port 1 Sep 21 23:02:19 Tower last message repeated 13 times Sep 21 23:02:19 Tower login[2327]: ROOT LOGIN on `pts/0' from `192.168.1.133' Sep 21 23:02:20 Tower kernel: hub 2-0:1.0: unable to enumerate USB device on port 1
September 22, 201015 yr Author I've finished running parity check, with 0 errors. Later I rebooted and regenerated syslog. I do not have any "additional" USB hubs connected to the box so am not sure what the original error message was all about. New syslog attached. I noticed that the same, or similar "error" messages are in this syslog as well. I am still not sure as to what the issue is with USB 1, but if I have nothing plugged into the PC other then the flash drive, should I be concerned and start looking for a new motherboard? I do not want the board to fail completely. Thanks, Lev syslog-2010-09-22_a.zip
September 23, 201015 yr Author Last night, with the unexpected storm, the power went out. I did not get a chance to power down unRaid box before the UPS died. When the power was restored, parity check was started and now I am getting a red ball next to disk1. ??? Any help? I've also included SMART log from what I think is that drive. I've run a short SMART test and am currently running the long test. I noticed that: "10 Spin_Retry_Count 0x0013 100 100 097 Pre-fail Always - 1" Does that mean that there was perhaps an error during the last boot that caused the drive to red-ball? syslog-2010-09-23_a.txt.zip smart.txt
September 23, 201015 yr a "red" ball indicates a "write" to the drive failed and it was taken out of service. Your smart report shows that the drive had to re-try spinning up once. That parameter is not yet failing (although a drive not spinning up indicates it is showing a LOT of wear OR the power supply is not up to the task) Since a write to that drive failed, you'll need to go through these steps to get it back online. Stop the array Un-assign the disk with the red-indicator on the devices page Start the array with the disk un-assigned Stop the array a second time. Re-assign the disk to its slot in the array. Start the array once more. It will re-construct the contents of the drive based on parity and the other disks onto the disk that had the red ball. It will have the full contents, including the "writes" that had failed to be written to the disk. Joe L.
September 23, 201015 yr Author Joe, Thank you for the reply. I was wondering... When the power went out last night, I tried to shut down the server to power it down, but as soon as I was able to log into it and issue "stop" command, the UPS died. I think that UPS went just as the disk1 was unmounting. Could that have caused it to try to spin up but fail in process as a result of power loss and not the intrinsic drive failure? Lev
September 23, 201015 yr Joe, Thank you for the reply. I was wondering... When the power went out last night, I tried to shut down the server to power it down, but as soon as I was able to log into it and issue "stop" command, the UPS died. I think that UPS went just as the disk1 was unmounting. Could that have caused it to try to spin up but fail in process as a result of power loss and not the intrinsic drive failure? Lev That could have been the cause. Based on the hours in the smart report, the drive is only about a year old. Depending on the UPS, it is possible to have unRAID automatically shut itself down in a power failure, before the UPS batteries are depleted. There is an "apcupsd" package that will work for all the APC brand UPS. (it is the one I use)
September 23, 201015 yr Author Joe, Thank you for the reply. I am rebuilding the drive at this point. Do you think that I should upgrade from 4.5.4 and if yes to what so that I can take advantage of the automatic power down? Lev
September 23, 201015 yr Joe, Thank you for the reply. I am rebuilding the drive at this point. Do you think that I should upgrade from 4.5.4 and if yes to what so that I can take advantage of the automatic power down? Lev You will also need to install unMenu and the APC package and the powerdown package to take advantage of the auto powerdown. If your UPS brand is not APC it may not work.
September 23, 201015 yr Author I am about 65% through the rebuild of the disk1 and noticed that one of the other data drives is spun down. Shouldn't all of the drives be accessed to reconstruct the data?
September 23, 201015 yr I am about 65% through the rebuild of the disk1 and noticed that one of the other data drives is spun down. Shouldn't all of the drives be accessed to reconstruct the data? It depends, if this spun down disk is smaller, once rebuild pass the size of this disk then there is no I/O goes to this disk when time is up. unRAID will spin it down even data rebuild is still in progress. However if this disk is at same size as the one you are rebuilding then something is not right.
September 23, 201015 yr Author You are absolutely correct. The drive which is spun down is 500GB compared to the 1.5 TB drive being rebuilt. Makes sense! Thanks! I am about 65% through the rebuild of the disk1 and noticed that one of the other data drives is spun down. Shouldn't all of the drives be accessed to reconstruct the data? It depends, if this spun down disk is smaller, once rebuild pass the size of this disk then there is no I/O goes to this disk when time is up. unRAID will spin it down even data rebuild is still in progress. However if this disk is at same size as the one you are rebuilding then something is not right.
Archived
This topic is now archived and is closed to further replies.