One of my servers has given a lot of trouble, most of which can be found in threads started by me here. Almost all of it was related to faulty hardware or the like.

 

For a long time now, I have missing files and folders in my shares. I access the server from several xbmc machines on the network. The machines are unable to locate many TV Show season folders. Only few of the folders show up. But if I browse the files from a windows computer on the network, or even by using the XBMC box's file manager I can see these "missing" files. Knowing the way XBMC scans media folders, its not possible for XBMC to overlook just a few seasons of TV Shows. To rule out the possibility of a faulty XBMC install I have tried many different machines. The shares are SMB.

I just realised that this server originally had a split level of 2 and after that I had once reinstalled unraid from scratch and at that time the split level was set to 1 though the disks and folder structure was the same. Because of this, entire seasons are missing and not episodes within season folders. However when I access the TV Shows share from Windows, everything appears as normal! How do I fix this issue?

My woes continue. Just when I have a nice array with all green balls, every now and then I return to find the server completely undiscoverable on the network!

 

I have no choice but to hard reset the machine. It comes on fine and starts a parity check. It works fine for the period of the parity check, but (I suspect) it crashes when the mover starts. I tried using MC and manually moving files from cache to the respective share, and sure enough it stopped abruptly at a file and from there onwards I have no contact with the server over the network.

 

I now plan to restart the server (by cutting power), disable cache drive and mover, let parity check complete and then worry about reconnecting the cache drive and starting mover etc.

 

Obviously the circumstances do not permit me to access a syslog

Network access is not required to get a syslog.  See SYSTEM LOG in my sig.

syslog attached.

 

Terminal closes few seconds after mover was manually started.

syslog.txt

run memtest overnight. See Check Disk Filesystems in my sig. Attach the entire syslog.

The "Check Disk Filesystems" link doesn't have any info on memtest. It refers to reiserfsck. You want me to perform that test or the memtest? Thanks!

Perform all tests to diagnose the problem.

I thought I should start by running reiserfsck disk by disk. On disk1 which I feel might be most likely to have errors, after a few mins of starting I get the following message:

 

reiserfsck --check started at Sun Sep  8 22:14:18 2013

###########

Replaying journal: Done.

Reiserfs journal '/dev/md1' in blocks [18..8211]: 0 transactions replayed

Killed

 

 

...and I am returned back to the command prompt.

 

What should I do next?

Need the entire syslog.

syslog attached

syslog.txt

Run memtest overnight. Then see my sig to disable all add-ons. Then try reiserfsck.

I have not got enough spare time on the server to be able to run the memory test (which I presume will put the server out of action for the period of the test)

 

I just thought I should add that I am able to copy from a Windows machine on the network, but on a particular disk it will not allow me to copy and gives the message "Destination Folder Access Denied - you need permission to complete this action". Also starting a reiserfsck on this disk gives a message at the very start saying that filesystem appears to be mounted read only.

 

I have tried running the new permissions utility but the situation is still the same. Will try to perform the memtest today.

The read-only disk requires reiserfsck to fix.

But the reiserfsck just gives a killed message and abruptly halts!

 

I ran a check and it is stuck for quite some time at this message..

 

Do you want to run this program?[N/Yes] (note need to type Yes if you do):Yes

###########

reiserfsck --check started at Thu Sep 19 21:54:53 2013

###########

Replaying journal: Trans replayed: mountid 169, transid 322054, desc 3328, len 1                                                                                        , commit 3330, next trans offset 3313

Trans replayed: mountid 169, transid 322055, desc 3331, len 1, commit 3333, next                                                                                        trans offset 3316

Trans replayed: mountid 169, transid 322056, desc 3334, len 1, commit 3336, next                                                                                        trans offset 3319

Trans replayed: mountid 169, transid 322057, desc 3337, len 8, commit 3346, next                                                                                        trans offset 3329

Trans replayed: mountid 169, transid 322058, desc 3347, len 21, commit 3369, nex                                                                                        t trans offset 3352

Replaying journal: Done.

Reiserfs journal '/dev/md2' in blocks [18..8211]: 5 transactions replayed

 

 

...and now I am back at command prompt.

 

 

FINAL EDIT:... and hooray! My shares work and are no longer read only.

 

 

Sorry but my joy was short lived. 

 

I managed to reverse the read only business by running reiserfsck but midway into a file copy to disk 2, I got an error message that the network location does not exist and I cannot access any shares nor web UI any more.

 

So it is pretty evident that when stuff is being written to disk 2 or shares contained on that disk some hell breaks loose on the array causing it to disappear from the network.

 

 

Post a SMART report.

SMART report for the disk in concern attached

smart.txt

I can't tell for sure but it seems like there isn't much amiss with the SMART report in the previous post. So I started a SMART long test by typing smartctl -t long /dev/sdc

 

It says it will take 255 minutes to complete. But where will I find the report for this?

same place.

Thank you for your help. I must have done something wrong, the only content of that smart.txt file was:

 

smartctl 5.40 2010-10-16 r3189 [i486-slackware-linux-gnu] (local build)

Copyright © 2002-10 by Bruce Allen, http://smartmontools.sourceforge.net

 

=== START OF OFFLINE IMMEDIATE AND SELF-TEST SECTION ===

Sending command: "Execute SMART Extended self-test routine immediately in off-line mode".

Drive command "Execute SMART Extended self-test routine immediately in off-line mode" successful.

Testing has begun.

Please wait 255 minutes for test to complete.

Test will complete after Sun Sep 22 20:55:38 2013

 

Use smartctl -X to abort test.

 

 

I can't figure out what went wrong.

 

Tell me, shall I just swap this disk out for another one, wait till rebuild finishes and check to see if the problem persists. Have we more or less isolated the problem to be a disk related one?

You'll need to run reiserfsck after the rebuild is complete. Then pre-clear the current drive.

I have changed the drive and the rebuild completed successfully.

 

But the problem is still the same - copying any data to disk2 causes the system to freeze. I lose network to the system. I assume the syslog also abruptly ends like the last one I posted.

 

So its definitely not the hard drive but there is something amiss about disk2 slot. Any suggestions?

Click on the small Log button on unRAID main in the upper right and leave the window open until it crashes.

Just as I had suspected, it crashes almost instantly. And the log also does not yield any info..

 

 

/usr/bin/tail -f /var/log/syslog

Oct 6 19:25:06 NAS emhttp: shcmd (49): /etc/rc.d/rc.nfsd start |& logger

Oct 6 19:25:06 NAS logger: Starting NFS server daemons:

Oct 6 19:25:06 NAS logger: /usr/sbin/exportfs -r

Oct 6 19:25:06 NAS logger: /usr/sbin/rpc.nfsd 8

Oct 6 19:25:06 NAS logger: /usr/sbin/rpc.mountd

Oct 6 19:25:06 NAS mountd[1530]: Kernel does not have pseudo root support.

Oct 6 19:25:06 NAS mountd[1530]: NFS v4 mounts will be disabled unless fsid=0

Oct 6 19:25:06 NAS mountd[1530]: is specfied in /etc/exports file.

Oct 6 19:25:06 NAS emhttp: shcmd (50): /usr/local/sbin/emhttp_event svcs_restarted

Oct 6 19:25:06 NAS emhttp_event: svcs_restarted

Oct 6 19:25:54 NAS emhttp: shcmd (51): /usr/local/sbin/mover |& logger &

Oct 6 19:25:54 NAS logger: mover started

Oct 6 19:25:54 NAS logger: moving Music/

Oct 6 19:25:55 NAS logger: ./Music/House/2013/Top 30 Dance Club Play 1.6.2013/19. Arianna ft. Pitbull - Sexy People.mp3

Oct 6 19:25:55 NAS logger: .d..t...... ./

Oct 6 19:25:55 NAS logger: .d..t...... Music/

Oct 6 19:25:55 NAS logger: .d..t...... Music/House/

Oct 6 19:25:55 NAS logger: .d..t...... Music/House/2013/

Oct 6 19:25:55 NAS logger: cd+++++++++ Music/House/2013/Top 30 Dance Club Play 1.6.2013/

Oct 6 19:25:55 NAS logger: >f+++++++++ Music/House/2013/Top 30 Dance Club Play 1.6.2013/19. Arianna ft. Pitbull - Sexy People.mp3

 

 

 

EDIT:

I just manually moved some files from cache drive to disk1 and that went without a hitch. So its definitely got something to do with disk2. I don't have any issues reading data from disk2, its just writing to it that seems to cause trouble. Could it be a bad motherboard SATA port? But I really doubt that. Just recently I reshuffled all the sata and power connectors on my rig and the problem still persists on disk2! Can't think what to do next.