My Hardware:

5x 2TB SAMSUNG_HD203WI (parity and disk1-4)

1x 1.5TB Seagate ST31500341AS (was running as cache, but with troubles "demoted" it to Disk5)

 

Asus M4A78LT-M LE Motherboard

AMD Sempron 140 processor.

2GB Dual Channel (two sticks 1GB) Kingston 1GB 1333MHz DDR3 Non-ECC CL9 DIMM KVR1333D3N9/

400W Corsair CMPSU-400CXUK PowerSupply

 

Background:

After some initial setup issues where the montherboard's interaction with my keyboard froze bootup (wanting to set NumCap on) I set it up and it seemed to work fine. I was able to copy over approx 4TB of DVDs and hundreds of thousands of work files, pictures etc into 4 shares.

 

Now with the extra space I thought I would combine a couple of my shares to allow my family to view pictures and DVDs from the same share....that's where the issues started.

 

My Problem - at least with unRaid :

Copying a directory of files (DVD for example) from my desktop to any of the shares works a treat, as you would expect. I can also browse and delete multiple files from the unRaid server without issues. I can even use a PopCornHour media server which works fine (using NFS).

 

The problem is when I try to read or copy any files back to my desktop or even another share on the same unRaid server I always get the same error message of: "Cannot copy NameOfFile: The specified network name is no longer available". At that point the file explorer window (\\videoraid\photo\etc) disappears, but I can go back to My Computer and immediately go to that directory again.

 

I dont get this error message immediately if I'm reading or copying a number of really small files (like desktop.ini) it will copy those before stalling and failing on any sizeable files it finds(my guess is 50KB or more). I have even built directory structures on unRaid using large numbers of 1KB files in each of 5 layers and it copies directories and files until it has copied approx 50-60KB of data at which point it then gives the same error message.

 

I note that when it is copying my test directory structure I see NO DRIVE ACTIVITY. So it is almost like the server isnt recognising the read request properly, to then access the disks. It seems to only be reading from the RAM cache. But when I explore the directory structure there doesn't seem to be any problem and I can see drive activity and more importantly the files and directories.

 

HELP!!

syslog.txt

I have a very similar problem. I have been helping a friend set up a completely new Unraid server. I have an older version running that has never given me any troubles.

 

The install went fine, drives formatted, parity check completed without errors. He can copy data to the server, but as soon as you try to open any of the files the server will lock up for about 5 minutes. You can see the shares on the network just fine, you just can't read any of the files. We originally thought it might be a problem with Windows 7 permissions, but we tried Windows XP and get the same problems.

 

We're running 3 x 2TB Seagate hard drives.

 

Hopefully these 2 problems are related.

After over 24 hours and no response from the Gurus of unRaid...I'm getting worried  :'(

I note that when it is copying my test directory structure I see NO DRIVE ACTIVITY. So it is almost like the server isnt recognising the read request properly, to then access the disks. It seems to only be reading from the RAM cache.

Linux works exactly like you describe. The disk buffer cache is accessed instead of the disks if the data is already in memory.  There is absolutely no need to go to the physical disks.

But when I explore the directory structure there doesn't seem to be any problem and I can see drive activity and more importantly the files and directories.

HELP!!

 

Step 1. post a syslog.  Instructions under troubleshooting in the wiki.  We can't help without the clues needed.

Nothing glaring wrong in the syslog...

I only see from 13:11:59 through 13:14:52,  a total of 3 minutes?  Does it cover the time-period where the file-copies failed?

I'm guessing it does not, since your trouble description says the copy proceeds for 5 minutes or more.

 

Step 2. Especially since this is a new server, perform a memory test.

 

Step 3.  Perform a file-system check on each of your data drives.  Instructions in the wiki.

http://lime-technology.com/wiki/index.php?title=Check_Disk_Filesystems

 

Step 4.  If the problem is only when copying from the Window's7 PC, then it is highly likely the problem IS the windows7 PC.  Try from a different PC if possible to isolate the cause.

 

Step 5.  It is as likely as anything that the router between the PC and the server is the issue.  Use a direct connection and see if the issue goes away.    Try a different cable...  One that is cat5e or cat6.

 

 

Joe,

 

Firstly let me thank you for helping out! This has really been driving me [move]nuts nuts nuts![/move]

 

So I will be so happy when we can solve it

 

Linux works exactly like you describe. The disk buffer cache is accessed instead of the disks if the data is already in memory.  There is absolutely no need to go to the physical disks.

I understand that but what I want to make clear is that it treats all read queries as a cache operation when clearly they are not! Also no matter where I am in the directories I can only read approx 50KB before it fails and times out.

 

I only see from 13:11:59 through 13:14:52,  a total of 3 minutes?  Does it cover the time-period where the file-copies failed?

I'm guessing it does not, since your trouble description says the copy proceeds for 5 minutes or more.

I think you are mixing dhilgendorf's and my issues. My timeouts take exactly 60 seconds. In any I created a new Syslog.txt file where I have done a fresh reboot, tried to copy/read twice and it failed on both occasions. This was between 16:35 to 16:47 on they Syslog.txt file and there is absolutely no errors showing.

 

Another observation is that when I tried to copy a HUGE directory structure 60,000+files it spent approx 4 minutes with "Preparing to Copy" showing on my desktop all the while it also had very active drive activity. It then started to copy to my desktop, immediately the drive activity stopped, 60 seconds passed and it timed out with an error.

 

Step 2. Especially since this is a new server, perform a memory test.

Did that both from the unRaid utility and also system bios option. Passed without issues in both cases.

 

Step 3.  Perform a file-system check on each of your data drives.  Instructions in the wiki.

http://lime-technology.com/wiki/index.php?title=Check_Disk_Filesystems

Did this on all 5 drives (not parity drive as warned in articles) used "reiserfsck --fix-fixable" but was unnecessary as all 5 passed without corruptions being detected.

 

Step 4.  If the problem is only when copying from the Window's7 PC, then it is highly likely the problem IS the windows7 PC.  Try from a different PC if possible to isolate the cause.

This is one where you are mixing the two different issues. I don't have Windows 7 on my network. But I do have Win Server 2003, Win XP and Vista. I have tried all three and all three have failed on reads but passed on writes. Also, you may not have noticed my original posting suggesting that a PopCornHour media box using NFS seems to be able to read flawlessly (It plays 9GB DVDs without issues). I think this is a SMB issue and not a hardware or NFS issue.

 

Step 5.  It is as likely as anything that the router between the PC and the server is the issue.  Use a direct connection and see if the issue goes away.    Try a different cable...  One that is cat5e or cat6.

I tried a direct connect using cat5e and there was no difference. But as above, because NFS appears to work (at least for a PopCornHour box on NFS reading DVDs) I doubt this is a hardware issue.

 

syslog.txt

Just a quick update on my problem. We solved it. It turned out to be a bad network controller on the server motherboard. Motherboard was brand new and the chipset was listed as compatible in these forums. Not really sure what the problem was. It worked fine for everything except reading from the server. Replaced it with a different network card and it worked instantly. I appreciate the help and am very happy to have this working now.

Just a quick update on my problem. We solved it. It turned out to be a bad network controller on the server motherboard. Motherboard was brand new and the chipset was listed as compatible in these forums. Not really sure what the problem was. It worked fine for everything except reading from the server. Replaced it with a different network card and it worked instantly. I appreciate the help and am very happy to have this working now.

 

Was yours working on NFS and reading large files?

Did it read the first 50k of files and then fail?

What motherboard/chipset where you using?

 

I'm just trying to establish if we possibly have/had the same problem.

 

Not sure about NFS. I don't really have any experience with that, so we didn't look into testing that too much. The motherboard was a GA-G31M-ES2L. It has a RTL 8111C chipset.

 

I'm not sure if it would read the first 50k then fail. It would definitely start to read a file, then the whole server would lockup for the next couple of minutes. I think I could still control it through the web interface. Sorry I can't be of much more help.

 

I originally tried a crossover cable to rule out the router and other cables. It still didn't work and luckily I had another network card sitting around.

Dhilgendorf, looks like your suggestion worked for me as well, thanks for posting that update!

 

I disabled the motherboard's ethernet port and installed a gigabit card and I can now read files

 

If anyone figures out why that may be or comes up with a solution please tell me because I would much rather use the motherboards ethernet rather than plug in card.

Having a similar problem - lately - it goes away after a reboot of the server.

Running 4.5.3 on a A7n8X running an Athlon XP chip at 2405Mhz

 

I think its time to upgrade my hardware - its so ancient.

Hi Guys,

 

I am having similar problems - I have tried with various flavours of windows (XP/2K3/7) with same results as well as trying to mount SMB shares on my Popcon A110 - I am running Unraid 4.5.6:

 

- I can upload via SMB without issue regardless of filesize type etc

 

- When trying to download via SMB I cannot even start the transfer (or playout on the A110). It pops up the window for 60ish sec, does nothing, then dies. I can play files out via the popcorn using NFS no problem but nothing can see the use the SMB share.

 

I see people have changed network cards and found the problem solved but when I tried this with a new GigE card I couldnt even get the web UI up.

 

So, my questions are as follows:

 

1. Is it hardware incompatability that is causing the problem. If so is there any way round it? Particular brand of NIC to use?

 

2. What more can be done to troubleshoot the problem?

 

3. Anyone managed to get NFS working on Win 7 - I have it properly enabled etc but it cannot map the Unraid NFS shares that the popcorn uses?

 

4. Has anyone tried using the SMB share in this condition with Linux machine - did that work (I will probably try this next...)?

 

All in all I am not having much fun right now and could do without spending my entire life solving this problem ;-)...

 

All help and suggestions greatfully recieved...!!

...changed network cards and found the problem solved but when I tried this with a new GigE card I couldnt even get the web UI up.

When you install the new NIC, you need to disable the old onboard NIC in the BIOS.

That will get your server up again.  Please report if your new NIC solves your problem or not.

 

 

Sorry I should have said that I did disable the onboard NIC via the BIOS - when the machine booted I had some issues with the default configuration of the new NIC - it was set to a strange multicast state and was not able to take an IP address. I did some reconfiguration with ifconfig and ethtool and got the NIC up properly, it then got an address via DHCP as expected.

 

Rebooted the machine again and all looked good however the UI never came up and I could not see any of the shares either. I then powered off, removed the card, re-enabled the onboard NIC and restarted again.

 

The onboard NIC came up as usual and I could access everything again - except the SMB shares of course...

 

Let me know if there is anything else I can provide - I have also checked the messaging etc (wireshark) and all looks OK there i.e. the devices are talking and passing messages, just not moving any data.

 

Many thanks!

Stumbled across this thread whilst researching the same problem as the OP.

 

I have the identical motherboard and Samsung hard drives as the OP

 

Adding a gigabit card and disabling the onboard LAN solved the problem for me too.

 

Cheers