January 27, 20179 yr I have three unRAID Plus servers running v5.0.5 One of the servers (which has been running fine for months) has just started being rather unstable. After a restart, it appears to run fine. However, the web GUI and shares will, after a couple of minutes, stop responding with the following message in Chrome: This site can’t be reached unraid-03 refused to connect. Search Google for unraid Main ERR_CONNECTION_REFUSED I also am unable to connect to shares via my Windows PC. I can ping it, and access it via the command line with a keyboard and screen attached. If I re-start it, I can again access the GUI, but again, after a very short period, it stops responding. Shares also. I was wondering whether an upgrade to v6.x would be wise, or whether I should perform some more diagnosis beforehand. I have a backup of the Flash drive, but only since the problem started occurring. I am not sure the file system is 100% clean, as I have had to power it down (reasonably gracefully) by using the power button to initiate a shutdown several times. It has not been written to recently, On booting the server, the web GUI looks like this: Once the server becomes unresponsive, after very many attempts to refresh it, I may be able to get the homepage to respond, at which time, the array appears to be offline: Many thanks in advance.
January 27, 20179 yr Author Thanks. I had to restart the server again to get access to the GUI to grab the Syslog from Utils. I've attached it as a TXT file, as it was apparently too big to cut & paste into the forum. unraidsyslog01.txt
January 27, 20179 yr Author Unfortunately not. I just tried booting into safe mode for the second time, and I immediately get the web GUI stop working on me.
January 27, 20179 yr Author I couldn't get FTP to work, but I finally realised that the latest logs were getting written to /logs, so I pulled out the Flash drive and stuck it in my PC to get the log file off. This is one which covers a frozen session (attached) - hopefully there is something useful in it, but I couldn't see anything obvious... You can see halfway down where I initiated a 'powerdown' at the command line following the freeze. I had to ZIP it, as it was too big to attach as a TXT file. syslog-20170127-203425.zip
January 27, 20179 yr Author It's very basic: #!/bin/bash # Start the Management Utility /usr/local/sbin/emhttp & The server is almost a default config - no packages running except APCUPSD. I hadn't made any changes to it for months. It just started playing up out of the blue. I was wondering whether maybe one of the drives was on the way out, or maybe the network card was about to blow up. Unfortunately my Linux knowledge is very basic, so I don't know how to test these things. I would be quite prepared to trash the server and install v6.x from scratch as long as I could re-mount the existing drive array. The only other option would be to buy a new set of drives, install those, and copy the contents of the old drives over one by one, by mounting them on my Windows PC. I'd rather not do that if I don't have to though, it's quite a costly option.
January 28, 20179 yr * Two of your SATA ports are still configured as IDE drives, and you connected your 4TB parity drive to one of those, which significantly restricts its performance, and therefore the write performance of your array. When you next boot, go into the BIOS settings for your motherboard, look for SATA settings, and make sure it's configured for AHCI. If AHCI isn't available, then select a native SATA mode, anything but the IDE emulation mode that it is in now. * Your flash drive may have some corruption. When you can, after shutting down, pull it and Check Disk it (may also be called ScanDisk) on a Windows computer. * This syslog was only 5 minutes. The server was booted, the array started, then 5 minutes later it was shut down. There are no issues evident at all, in that short time. What would be useful is a syslog captured *after* things fail, if that's possible. We want to see the errors that are logged. * Zipped syslogs are what we like.
January 29, 20179 yr Author Hello. Many thanks for your assistance so far. I ran a CHKDSK on the Flash drive, and it came up as OK - no issues. The log I posted before was actually covering the whole incident from boot to freeze to shutdown. I can get it to stop responding almost instantly. ---- As an aside, I was wondering whether I can copy my files off the array using another machine rather than trying to repair it in-situ. I have a laptop running Ubuntu 16.04, and a USB SATA dock plugged into it. If I pull one of the 4 drives from my unRAID array and connect it up to my Ubuntu machine using the USB SATA dock, will I be able to read the files? Will doing so stuff up the drive so that the array will no longer work? (Even if I don't write to it)? Many thanks.
January 29, 20179 yr As an aside, I was wondering whether I can copy my files off the array using another machine rather than trying to repair it in-situ. I have a laptop running Ubuntu 16.04, and a USB SATA dock plugged into it. If I pull one of the 4 drives from my unRAID array and connect it up to my Ubuntu machine using the USB SATA dock, will I be able to read the files? Will doing so stuff up the drive so that the array will no longer work? (Even if I don't write to it)? Yes, your Ubuntu should have no trouble reading the drive, should auto recognize its ReiserFS 3.6 file system. If at all possible, mount the drive read-only, and you shouldn't have any problems breaking parity, if you put it back.
January 31, 20179 yr Author Indeed, I had no trouble with the Ubuntu server reading the drives. Still copying files off them, and will rebuild as v6 before coping them back over. Many thanks.
February 7, 20179 yr Author Once I backed up the files to my NAS, I formatted the Flash drive and installed version 6.2.4 from scratch with default settings, and copied over my v5.x license file. Booting up as a default, blank install, it listed all my existing drives from the v5.x array. I selected the drives into the array in the same order they were before, and the parity drive started to rebuild (this is taking a long time, as there is 14TB of files across the 4 disks). The original shares appear to be intact, and I can see all the files. All that is missing are the original user accounts. So it appears at this time that all I need to do is re-name the server, create the old accounts, set up the UPS settings and I'm finished. No need to copy all the files back over. Unless I'm missing something... I had custom split levels set for each share on v5.x - I'm not sure how the automatic split method is going to work on v6.x, but I guess I'll see...
February 7, 20179 yr Once I backed up the files to my NAS, I formatted the Flash drive and installed version 6.2.4 from scratch with default settings, and copied over my v5.x license file. Booting up as a default, blank install, it listed all my existing drives from the v5.x array. I selected the drives into the array in the same order they were before, and the parity drive started to rebuild (this is taking a long time, as there is 14TB of files across the 4 disks). The original shares appear to be intact, and I can see all the files. All that is missing are the original user accounts. So it appears at this time that all I need to do is re-name the server, create the old accounts, set up the UPS settings and I'm finished. No need to copy all the files back over. Unless I'm missing something... I had custom split levels set for each share on v5.x - I'm not sure how the automatic split method is going to work on v6.x, but I guess I'll see... There should have been a check box you could have used to tell it parity was already valid so it wouldn't have to sync parity again. And the amount of time parity takes is unrelated to the total capacity. It is mostly just a function of how large the parity disk is, disk speeds, and controllers. unRAID automatically creates shares for all the top level folders on cache or array disks, but they will have default settings until you change their settings. If you made a backup of your flash you could copy the share settings from your flash backup in config/shares folder. Some of your other settings, including users, are also in the config folder of your flash backup. All of the .cfg files on flash are just text and you can probably examine them and make some sense out of them yourself. You might take a look at the Upgrade link in my sig for some more insight into what you can copy from config if you did make a flash backup.
February 9, 20179 yr Author What's strange is that, once the parity drive re-built, the shares all disappeared from the GUI. I know they're there, because I can connect to them all remotely. I can also add new shares, but once they're added, they don't appear in the GUI either. I'm guessing there's some error which is stopping the GUI from pulling a list of them. Which is a bit annoying.
Archived
This topic is now archived and is closed to further replies.