September 4, 201015 yr Hi all...I replaced a 250gb disk with a 750gb disk today. I restarted the array and it said it would take about 240min. to reconstruct. I left it for more than that time and then pressed refresh and it wouldn't connect to the server. Any ideas? THANKS! UPDATE: I just got home tonight, I still can't connect via the web browser. I got in with telnet and got a syslog. I can also access the disks, just nothing from the browser. I'm including the syslog here. I'm not sure what I'm seeing, but it's late and I'm pretty fried. Will look again tomorrow, but if anyone can look at this and tell me if they see anything, I'd much appreciate it. 9_4_10_syslog.txt
September 5, 201015 yr Author UPDATE: Since I got no response, I tried the directions in the following thread: http://lime-technology.com/forum/index.php?topic=2419.msg18821#msg18821 I got back some unexpected responses: root@unRAID:~# /root/samba stop root@unRAID:~# for i in /dev/md* > do > killall 'fuser -c $i' > unfount $i > done fuser -c $i: no process killed -bash: unfount: command not found fuser -c $i: no process killed -bash: unfount: command not found fuser -c $i: no process killed -bash: unfount: command not found fuser -c $i: no process killed -bash: unfount: command not found fuser -c $i: no process killed -bash: unfount: command not found fuser -c $i: no process killed -bash: unfount: command not found fuser -c $i: no process killed -bash: unfount: command not found fuser -c $i: no process killed -bash: unfount: command not found root@unRAID:~# /root/mdcmd stop cmdOper=stop cmdResult=failed root@unRAID:~# reboot Broadcast message from root (pts/0) (Sun Sep 5 13:19:18 2010): The system is going down for reboot NOW! The server dropped offline, but it didn't restart. So, I had to do a hard shutdown. When I restarted, it started rebuilding the new disk again, so we'll see after 5 hours or so and see what happens. Here's the last syslog before I shut down: 9_5_10_syslog.txt
September 5, 201015 yr UPDATE: Since I got no response, I tried the directions in the following thread: http://lime-technology.com/forum/index.php?topic=2419.msg18821#msg18821 I got back some unexpected responses: root@unRAID:~# /root/samba stop root@unRAID:~# for i in /dev/md* > do > killall 'fuser -c $i' > unfount $i > done fuser -c $i: no process killed -bash: unfount: command not found fuser -c $i: no process killed -bash: unfount: command not found fuser -c $i: no process killed -bash: unfount: command not found fuser -c $i: no process killed -bash: unfount: command not found fuser -c $i: no process killed -bash: unfount: command not found fuser -c $i: no process killed -bash: unfount: command not found fuser -c $i: no process killed -bash: unfount: command not found fuser -c $i: no process killed -bash: unfount: command not found root@unRAID:~# /root/mdcmd stop cmdOper=stop cmdResult=failed root@unRAID:~# reboot Broadcast message from root (pts/0) (Sun Sep 5 13:19:18 2010): The system is going down for reboot NOW! The server dropped offline, but it didn't restart. So, I had to do a hard shutdown. When I restarted, it started rebuilding the new disk again, so we'll see after 5 hours or so and see what happens. Here's the last syslog before I shut down: The command was "umount" , not "unfount" Joe L.
September 6, 201015 yr Author oops OK, tried it again and got: root@unRAID:~# /root/samba stop root@unRAID:~# for i in /dev/md* > do > killall 'fuser -c $i' > umount $i > done fuser -c $i: no process killed umount: /dev/md1: not mounted fuser -c $i: no process killed umount: /dev/md2: not mounted fuser -c $i: no process killed umount: /dev/md3: not mounted fuser -c $i: no process killed umount: /dev/md4: not mounted fuser -c $i: no process killed umount: /dev/md5: not mounted fuser -c $i: no process killed umount: /dev/md6: not mounted umount: /dev/md6: not mounted fuser -c $i: no process killed /root/mdcmd stop Broadcast message from root (pts/2) (Sun Sep 5 19:17:20 2010): The system is going down for reboot NOW! I'm trying the tail -f /var/log/syslog command to track the syslog and will upload it if it freezes again. Thanks Joe.
September 6, 201015 yr Author Hmm, this is getting interesting. I've tried several things. I removed the new 750gb disk from the 5-3 cage with backplates, thinking something might be up with that, but got the same problem with a direct connection. Then I swapped out the sata cable and tried a different power cable. No luck. Then I removed the 750gb disk all together and the syslog looked good and normal (nothing red in unmenu's syslog), but of course now unRAID is reporting a missing disk. Finally, I tried a different 750gb disk and I'm getting the same problem! So now I'm wondering if the sata card is crapping out. I'm am including the latest syslog. I think I'm going to put back the old 250gb disk and see what happens. I assume it will want to rebuild that one too because it's forgotten that one, right? One thing is bothering me, though I don't see how this could affect what's going on. Although the 250gb drive that I'm trying to replace was part of the array, I had no files on it. Could that be causing something to go wonky? THANKS!!!!! 9_5_10_syslog4.txt
September 6, 201015 yr nothing red in unmenu's syslogunMENU colors a line "red" if it has the word "error" somewhere in it. It is not the definitive analysis you might need. Although the 250gb drive that I'm trying to replace was part of the array, I had no files on it. Could that be causing something to go wonky? No.
September 6, 201015 yr Author I left it on overnight and I can access all the disks as normal shares, I can reach the server through telnet, but I still can't get there with the browser. Can anyone tell from these syslogs what's going on? I'm including the last one and the main thing I see is it has continual trouble with ata5 and ata6. I'm not well versed enough to know what that means. Are these hosts? I have two SATA cards in there, could this be showing that one has gone bad? But meanwhile I have access to everything through the finder (mac). Also, if I try to shut the server down through telnet or the power button, it balks and I have to hold the power button until it shuts down. I know that's not good. My next thought is to swap my two SATA cards and see if the problem follows the card. If not, I'll try to put back the old 250GB drive and see what happens. Does that sound right? 9_6_10_syslog.txt
September 7, 201015 yr Author I think I've run into an ongoing problem that didn't really show itself until now. I looked back on some of my older posts and syslogs and noticed that ata5 and ata6 have been having trouble, throwing errors, stepping down in speed. Now ata6 has been dropping down to UDMA/33 and then this freezing. I also was having some problems in the past with accessing the server via http. I think it's the SATA card. I just tried swapping the two SATA cards, but then 4 of the disks showed up in the wrong place. I thought the disks would be associated with the PCI slot, not the card, but that isn't the case. Once I worked that out, changing cables, etc, things got even worse. The rebuild slowed down even more and the system would freeze at even earlier points in the rebuild. I wonder if that's because one of the slots is a straight PCI and the other is a PCIe-16x. Anyway, next, I tried to put back the original 250Gb drive, but now the array was expecting a 750Gb drive and said the 250 was too small. I was tempted to try the initconfig command, but I really don't think this has anything to do with the drives, and I'm afraid it would complicate things even further. I think my next move should be to get some better cards and see what happens. Any suggestions for good low cost SATA cards? I have one PCI slot and one PCIe-16x slot available. Do you think I'm on the right track? I also have a basic question, I'm getting confused with all the disk and host identifiers. There's disk(x), ata(x), md(x), sd(x) and host(x). Which is which? THANKS!
October 9, 201015 yr Author If anyone is interested, I solved the problem. It was a bad SATA card. The server is working perfectly now.
Archived
This topic is now archived and is closed to further replies.