Hello.

 

I am very new, and you guys probably have heard many of my problems before.  I apologize in advance. I'm am both new to Unraid, and new to Linux environments.

 

I recently purchased a AVS-10/4 Server.  I had it preconfigured with 6 drives (parity plus 5).  Also it has 2x SSD cache drives, 32gb of ram.  I am not currently running any additional applications with it.  The only modifications that I made to the were ones in the Configuration Tutorial, specifically installing Unmenu and Screen, Configuring Mail and SSMTP.

 

I set my shares up as described in the Tutorial.  Then I transfered 12TB of movies/TV shows over.

 

I was ectatic.  Everything was working perfectly.  Very fast.  Responsive.  I was bragging to my friends.

 

Then, I probably screwed something up.

 

I took 3 older 4TB seagate drives (same models that were preconfigured by LimeTech when I ordered, ST4000DM000), and I ran 3 sessions of the preclear script on each drive.  I received no errors, and then added them into the array.  Still everything seemed to be working well.

 

Somewhere in here, occasionally the WebGUI becomes unresponsive (though it seems I can still Putty/Telnet in).  Also, occasionally, my shares seem to randomly go offline.  One additional error was that when transfering a few movie directories (about 50gb of files) to the media share (which has about 15 TB free), it says there is no space available for the transfer.

 

Since I'm so new, I'm not sure where to start and was hoping someone could point me in the right direction.

 

-Did I screw something up?  I added the drives according to the tutorial above

-Did I mis configure my shares somehow? since the SSDs run as cache, maybe they were full, and so I had not enough space there.  But why wouldn't it just right to the platter drives when the SSD was full?

Other?

 

Where can I start? I feel underqualified for this. Are there files that I can attach to this post that would help?

Hi,

 

Yes, you can attach your system log (syslog). It is located in /var/log.

 

As long as you can telnet into your system just copy it to your array:

 

cd /var/log

cp syslog /mnt/disk1

 

then from your windows server get it from your disk1 share

 

(or cp it to a user share: cp syslog /mnt/user/<nameofyourshare> )

 

Also please mention the unraid version you are using.

 

For your own troubleshooting process: disable all your plugins (if any) and try again, mostly these issues are plugin related..

 

With respect to the error when transferring files: could it be that your cache drive was full ?  If you are using the cache drive then the files are not actually copied to your drives that have 15tb free.. but to your cache drive..

So if the cache drive (512gb) is full during a copy, it won't simply copy directly to the regular drives?  Is that something that can be changed?

 

The telnet connection was working, though stopped before I could try to download the syslog.  A reset was done, and I have the syslog.  From what I've read, this is a different syslog. Sorry.

syslog-2015-03-31.txt

I must have really screwed something up here.  I just tried to copy to my media share.

 

-It takes a very long time to display the contents

-The transfer (from within windows via SMB shares (mapped)) fails, it did not even attempt it.  Just hangs when I copy and paste.

 

The WebGUI is still working, so I copied syslog from the UnMenu syslog page, I've attached the new one below.

 

I'm so sad.

 

-Edit: does it matter that a parity check is still going from the previous failure when I try to copy the new files.  Can you add files during a parity check?

syslog-2015-03-31.txt

So if the cache drive (512gb) is full during a copy, it won't simply copy directly to the regular drives?  Is that something that can be changed?

 

The telnet connection was working, though stopped before I could try to download the syslog.  A reset was done, and I have the syslog.  From what I've read, this is a different syslog. Sorry.

 

Yes and no. It depends how you set it up. As long as you set a "minimum free space" large enough it will go directly to the array if that minimum space is reached..

 

The following thread describes it in a bit more detail:

 

http://lime-technology.com/forum/index.php?topic=16715.0

 

I must have really screwed something up here.  I just tried to copy to my media share.

 

-It takes a very long time to display the contents

-The transfer (from within windows via SMB shares (mapped)) fails, it did not even attempt it.  Just hangs when I copy and paste.

 

The WebGUI is still working, so I copied syslog from the UnMenu syslog page, I've attached the new one below.

 

I'm so sad.

 

-Edit: does it matter that a parity check is still going from the previous failure when I try to copy the new files.  Can you add files during a parity check?

 

You can do everything you want with the array when a parity check is running but writing will be sloww... I usually try to avoid doing much with the array when a check is running.. But you cannot break anything.

Your syslog is showing errors on your ATA1 interface. Not looking good as far as I can tell. Check your internal cabling and make sure no connectors are loose.. Also check your SMART values (click on disk, then HEALTH, then disk attributes), check if your have reallocated sectors/offline uncorrectable/pending sectors and tell us the numbers behind them..

 

It might be worthwhile to check which disk is behind that ATA1.00 interface.

 

If you run

dmesg | grep ATA1.

then it should tell you which disk it is..

 

Unraid gives you a red ball (red ball before the drivenumber) when it detects a failed write, if you do not see this it does not mean nothing is wrong.. You should check this out..

In the syslog, it was referencing this:

Mar 31 17:07:52 Orcrist kernel: ata1.00: ATA-9: ST4000DM000-1F2168,             Z302NTZJ, CC54, max UDMA/133 (Drive related)

I spotted those error lines on ATA1.00.  Does the line mean that this drive is the ATA1.00 drive, which is serial number Z302NTZJ?  Z302NTZJ is my Parity drive.  I was wondering if the error references were to the Parity drive since it had to do a parity check after a reboot. Would that be a normal error message?

root@Orcrist:~# dmesg | grep ata1
ata1: SATA max UDMA/133 abar m2048@0xf7512000 port 0xf7512100 irq 36
ata1: SATA link up 3.0 Gbps (SStatus 123 SControl 300)
ata1.00: ATA-9: ST4000DM000-1F2168,             Z302NTZJ, CC54, max UDMA/133
ata1.00: 7814037168 sectors, multi 16: LBA48 NCQ (depth 31/32), AA
ata1.00: configured for UDMA/133

It looks like those 2 things do correlate, since the command above gave the same drive: Z302NTZJ.

 

It's worth mentioning that after the parity check, I rebooted the server, and there are no error messages in the syslog anymore.  I did take copies of multiple syslogs prior to the reboot.  So could those errors have just been references to the Parity error, since the drive referenced was the parity drive?

Yes, it is referencing your 4tb seagate drive..

 

If there are no errors anymore then that is a good thing, but do check cabling and your SMART values for this drive.

 

The errors would not come from the parity check