I build a small array (a 3TB data drive and a 4TB parity drive) to test out unraid. If it works for what I want I plan to get the pro key and add the rest of my drives, 9 more. I ran preclear on both drives a few times and have the array up and running. The parity check went without issue. I'm having an issue that I can't copy to the array from my windows 7 machine... well some files do get copied before the error occurs. I've tried both teracopy and the standard windows copy and both have the same result, and both appear to be loss of network connection. weird thing is though is there is no issue with connectivity and the web based interface communicates with the server just fine. I've done some searching for the "Specified Network No Longer Available" error that teracopy gives me and found quite a few similar threads and all of them go stale before any fix is found. They also appear to happen between 2010 and 2012 so they're not new threads by any means. I'm hoping a solution to this issue has since been found. I did notice that quite a few people were getting around this issue by using a cache drive.

 

I've attached my syslog file and a screenie of the teracopy failure, also network info in case it might help.

 

Thanks for any help!

Ryan

syslog.txt

ifconfig.txt

ethtool.txt

The syslog does not include the period of the failed transfers. Provide a syslog that coincides with a failure.

The syslog does cover the time when the errors occur... actually about 4 or 5 failed attempts. Nothing gets written to the syslog when it errors for this.

Try rebooting the router and all switches that you have connected to your network.  (All of these are really computers and sometimes they need a reboot as a result of old term software/hardware issues.)

Teracopy indicates 7:43 PM. The syslog starts at 20:50. These times are not coincidental.

I made the screen shot of the teracopy earlier, before I started this thread I tried a few things first with no luck. I didn't bother making a fresh screen shot of the failed copies as the error message it the same.

Use a fixed ip for the unraid?

 

That might help.

 

You can set it up from the webGUI.

I have some success! I tried all the above suggestions with no luck. After trying a few more things I decided to try upgrading from 5.0.1 to 5.0.5 by copying/overwriting the bzroot and bzimage files on the flash drive and all seems to be going well. Not sure if it was a borked initial install or something was fixed from 5.0.1 to 5.0.5 to work with my hardware but it's happily copying away now.

Thanks for your efforts.

Glad you figured it out. There were some 'small' changes from 5.0.1->5 that fixed rare hardware specific errors, as well as a few feature 'tweaks'.

Looks like I got a little ahead of my self here. I'm having the same problem again. I did end up getting the first of my media drives copied over to the unraid server but starting on the second disk this error has surfaced again. It did show a couple times when copying the first disk bot I got it done by only moving a few movies at a time. I tried all the stuff I did before with no success. One thing I didn't mention (got it fixed before I could post this part) Was when I try to delete the failed copy I cant because it says the file is in use. I've attached a screenie of that as well as a syslog. If I reboot my windows machine this error will still persist, if I take the array off line for a moment and restart it then I can delete the failed copy. Also this only seems to happen on larger files ~1GB and larger. I tried having the source drive in the windows machine as well as in the unraid machine using SNAP.

 

Any ideas?

syslog.zip

Are you transferring over wifi?

No, wired.

More clues...

 

Not sure if this a cause or effect but I tried runnng the 'fuser' command and when operating normally I only see one user running the command smbd. When the copy fails a second user shows up - this is when I can't delete the failed copy because I get an error that the file is open by another program. When I run the 'kill' command and select the PID that has the capital F under the ACCESS header, I can then delete the file. Attached is a screen shot. Not sure what to make of this.

 

I should add the 1st instance I ran fuser (in the screen shot) was before I started copying.

The 2nd time was during the copy while copying normally.

The 3rd is when the copy failed and teracopy was timing out.

the 4th is after I cancelled the teracopy. At this point I couldn't delete the failed copy.

After I killed PID 5091 I was able to delete the failed copy.

Bump.

 

Sometimes it works, sometimes it doesn't. Just wondering if anybody has any new ideas.