March 7, 201115 yr Hey everyone - I just built my first UNRAID server, and I'm having a few issues I'm hoping you can help me with. First, let's get the obvious out of the way. This is new territory to me, and I did some things I shouldn't have done. I'm well aware of this. Ok, so I built the box and managed to get it up and running with UNRAID 5.0-beta6. I'm currently running the free license, with the intention of upgrading to pro once everything's going swimmingly. I have four drives connected to the array (I thought the free server could do 6 tb of storage rather than 6 tb total), so one drive isn't doing anything at the moment. I precleared all 4 drives. Once that was done, I added three of them to the array. Everything seemed to be going fine for a day or two, and then this morning I started having some trouble accessing some of my user shares. I could still access everything through the disks (I'm running UNRAID on Mac OS X, so I see two server access points. One is called UNRAID, which houses all of the user shares. The other is called UNRAID-SMB, which lists the hard drives in the array), but certain user shares were giving me trouble. I tried unmounting them through the web UI, thinking a remount might clear things up. It got caught up in an endless cycle of waiting for access points to close or something similar. So, foolishly, I went to the server box and powered things down. When I powered back up, I couldn't access anything, and a reiserfsck tells me I have bad blocks on disk 1. When I go to the webUI, it shows the parity disk mounted, but has nothing selected for disk1 or disk2. The only option for either disk is the fourth hard drive which was previously not in the array. What are my options at this point?
March 7, 201115 yr Version 5b6 is full of bugs. It is intended for testing only. Use version 4.7 with NFS shares until version 5 is stable.
March 7, 201115 yr Author Thanks for the suggestions. I have downgraded to 4.7, but am still having issues. Next to parity disk, it now says "Missing _3D", and it simply says "Missing" next to disk1 and disk2. In the command area, it says "Stopped. Invalid Configuration." and "Too many wrong and/or missing disks!" It doesn't seem to want to recognize any disks now. I've attached a syslog file to this post. syslog.txt
March 7, 201115 yr Mar 7 17:10:55 UNRAID kernel: ata3.00: link is slow to respond, please be patient (ready=0) Mar 7 17:10:55 UNRAID kernel: ata3.00: SRST failed (errno=-16) Mar 7 17:10:55 UNRAID kernel: ata3.00: link is slow to respond, please be patient (ready=0) Mar 7 17:10:55 UNRAID kernel: ata3.00: SRST failed (errno=-16) Mar 7 17:10:55 UNRAID kernel: ata3.00: link is slow to respond, please be patient (ready=0) Mar 7 17:10:55 UNRAID kernel: ata3.00: SRST failed (errno=-16) Mar 7 17:10:55 UNRAID kernel: ata3.00: limiting SATA link speed to 1.5 Gbps Mar 7 17:10:55 UNRAID kernel: ata3.01: limiting SATA link speed to 1.5 Gbps Mar 7 17:10:55 UNRAID kernel: ata3.00: SRST failed (errno=-16) Mar 7 17:10:55 UNRAID kernel: ata3.00: reset failed, giving up First, check all cables and connections, one may be loose. Do these drives actually contain data that you care about? Do we need to worry about data loss here? Or is this a test server where we can proceed with reckless abandon?
March 8, 201115 yr Author I'll check that as soon as I can, but would one loose connection cause the errors with the other 3? As for your question, there is a small amount of data on the drive, but nothing irreplaceable. I had only just begun copying things over to the drive, and everything is backed up so we can proceed with reckless abandon. Also, if it makes a difference, all of the drives were fresh out of the packaging, never used until they entered the array.
March 8, 201115 yr I'll check that as soon as I can, but would one loose connection cause the errors with the other 3? If the loose connection is to a power splitter powering multiple drives, then certainly. If the loose connection is to a single drive, then probably not. Maybe there's no loose connections at all, but it is still worth checking. As for your question, there is a small amount of data on the drive, but nothing irreplaceable. I had only just begun copying things over to the drive, and everything is backed up so we can proceed with reckless abandon. Glad to hear that. If we don't find anything hardware-based, then the next steps I'll recommend will put any data on the array at risk of a disk failure (you'll temporarily lose parity protection). Also, if it makes a difference, all of the drives were fresh out of the packaging, never used until they entered the array. You said they all passed preclear, so I'll assume the drives themselves are good.
March 8, 201115 yr Author Ok, just got home and checked everything out. There didn't seem to be any loose connections, although the server did start up considerably faster this time around. I'm still not having any drives recognized though.
March 8, 201115 yr Post another syslog? Look for drive errors like in Rajahal's post. Did you go to the devices page and check that you could assign the drives? Peter
March 9, 201115 yr Author A new syslog is attached. I'm not receiving the same errors that were pointed out, but the drives continue to not be recognized. I have also attached a screenshot of my web interface, where the drives don't show up. syslog2.txt
March 9, 201115 yr All 4 disks appear to be connected Mar 8 17:45:27 UNRAID emhttp: pci-0000:00:1f.2-scsi-0:0:0:0 host3 (sda) WDC_WD20EARS-00MVWB0_WD-WCAZA1221197 Mar 8 17:45:27 UNRAID emhttp: pci-0000:00:1f.2-scsi-0:0:1:0 host3 (sdb) WDC_WD20EARS-00MVWB0_WD-WCAZA1305209 Mar 8 17:45:27 UNRAID emhttp: pci-0000:00:1f.2-scsi-1:0:1:0 host4 (sdc) Hitachi_HDS5C3020ALA632_ML0221F304RL3D Mar 8 17:45:27 UNRAID emhttp: pci-0000:00:1f.5-scsi-0:0:0:0 host5 (sdd) Hitachi_HDS5C3020ALA632_ML0221F3046ZUD I will ask again - did you go to the devices page (look near the top just below and to the right of the Lime Technology) and try to re-assign the disks? FYI - the main page will never have disk selection drop-downs like 5.0b6 had. Peter
March 9, 201115 yr Author FYI - the main page will never have disk selection drop-downs like 5.0b6 had. Well I feel like a goon. I was looking for the 5.0b6 interface. I now have the disks assigned, but it still says "Too many wrong and/or missing disks." I tried a few different combinations of assigning the drives, but continued to get the error. Here is a screenshot of the current interface. Should I continue to post syslogs? I don't think anything's really changing on the server side of things.
March 9, 201115 yr FYI - the main page will never have disk selection drop-downs like 5.0b6 had. Well I feel like a goon. I was looking for the 5.0b6 interface. I now have the disks assigned, but it still says "Too many wrong and/or missing disks." I tried a few different combinations of assigning the drives, but continued to get the error. Here is a screenshot of the current interface. Should I continue to post syslogs? I don't think anything's really changing on the server side of things. You'll probably need to set a new initial configuration. The old config files and the new (in the 5.0b6 version) are completely different in format.
March 9, 201115 yr Ya, log on at the terminal or via telnet and type initconfig. Answer with Yes. Refresh the web interface and you should be good to start. You could also type -> //"your_server_name"/blank.htm?cmdInit=apply in a web browser but replace "your_server_name" with Tower or whatever you called the server. This will reset with no confirmation prompt so be careful using it and that you don't accidentally use it at a later date due to say it being kept in your browser address pull-down. You could also just erase the super.dat file on the flash drive. Share the flash, go to it, delete super.dat and then refresh the browser window. Peter
March 9, 201115 yr Author That appears to have done the trick, at least for now. Thank you all so much - your patience with my inexperience is very much appreciated.
Archived
This topic is now archived and is closed to further replies.