I have posted a message about this issue before but go no repsonse.  Trying here again......

 

Nearly 95% of the time, and 100% of the time if I have not used the system in a while....attempts to move large files from my Windows 7 system to the Unraid will fail on the first attempt.

 

When I move a file - I get "Moving" status window up.  If I click on More Details - everything reports as "Calculating.  (For Time Remaining, Files Remaining, and Speed).  After 4-5 minutes it will fail.  I have to click Cancel and wait another minute while the system refreshes.

 

If I then try to move the exact same file again to the exact same location - it works without any issues.  Transfer Speeds up around 25-30mb/sec

 

This is exteremely frustrating and I can't figure out why this is happening.  At first i thought it was because maybe the drives were "asleep" but that does not seem to matter.  When I disable the sleep feature for the drives the same errors occur.

 

Anyone with any insight?

I use Windows 7 also, and I've never had this problem.  What are the details of your system and network?  Workgroup defined correctly?  Static IP or DHCP?  Anything else on your network that could be causing interference?  Is the unRAID server the master browser?  Are you using SMB and only SMB?

 

Also, how are you accessing unRAID from Windows 7?  Mapped drives?  Shortcuts?  Browsing through 'Network' each time?

 

I agree that it sounds like a drive spin down problem, but you say you've already tested that.  Still, I would like you to perform one more test:

 

On the unRAID web management page, either click the 'sync file systems and spin down' buttons or wait for some disks to spin down.  Once your disks are spun down, click the 'spin up' button to spin up all your disks.  Now attempt a transfer.  Does it work?

 

Finally, post a syslog.  Couldn't hurt.

I would also suggest getting TeraCopy which will give you a more real-time view of what's happening when you are copying files. This helped me when I was troubleshooting. http://www.codesector.com/teracopy.php

I second that, TeraCopy is great.  I actually meant to write that in my post, but forgot.

Also, something I ran into with Windows 7 awhile back that had been driving me nuts is it apparently trying to Power Manage the NIC.

 

I use mapped network drives and I would loose connection to them maybe once or twice a week. Other than rebooting I figured out I could restart the Workstation service to get them back. Then I read a post about the Power Management thing. In the properties for the NIC is a Power Management tab where you can disable this, which I did. I haven't had a problem since.

Nearly 95% of the time, and 100% of the time if I have not used the system in a while....attempts to move large files from my Windows 7 system to the Unraid will fail on the first attempt.

 

 

I believe you are describing a symptom that happen on my system as well. This is not show stopper but annoying

 

I also use Windows 7 and have multiple user shares mounted in this Windows 7 machine as network drive. If I shutdown my unRAID for multiple days, all those mounted network drives will become unavailable that is correct. However once I boot up my unRAID, all those network drives after I access them will resume to online except one of them. This particular user-share-turn-to-network-drive will have problem first time I access it. From Windows when I copying data to this particular drive I can see Windows is trying to refresh information of this drive because a progress bar is going on but for some reason it never finish and eventually timeout. After timeout if I do it again then at 2nd time no issue at all in copying data over and never will have. Teracopy is not going to solve it because I am using it.

 

Because this symptom only happen on this particular user share so I tend to believe there is some issue in unRAID in determining where to put those new data. Because a user share might across multiple disk and to determine how/where to place data unRAID will need to find out which disk has more space….etc and from unRAID WebGUI I do see read counts for those disks involved in this particular user share increasing. In the mean time, if I use disk share I never have this problem. So far, however I still can’t pinpoint exactly which module to blame.

 

 

 

Having the exact same issues on a new build of unraid with 2x2TB EARS (one parity), copying from a Windows7 PC. The server hardware was a Windows Home Server before and it worked perfectly fine. Samba seems to stop responding then timeout and then finally respond again. For example, I copied my syslog to a txt file on the array to access via the windows PC. After double clicking on the txt file notepad locked for ~30 seconds before giving the error of the network name is no longer available.

 

I have modified my /etc/samba/smb.conf with the following:

        log level = 2

        syslog = 2

 

And restarted the daemon, here's the log:

 

Nov 11 03:50:23 Magi smbd[3352]: [2010/11/11 03:50:23,  1] smbd/service.c:1047(make_connection_snum)

Nov 11 03:50:23 Magi smbd[3352]:  pc (192.168.11.117) connect to service Testshare initially as user root (uid=0, gid=0) (pid 3352)

Nov 11 03:50:28 Magi smbd[3352]: [2010/11/11 03:50:28,  0] smbd/nttrans.c:2119(call_nt_transact_ioctl)

Nov 11 03:50:28 Magi smbd[3352]:  call_nt_transact_ioctl(0x900eb): Currently not implemented.

Nov 11 03:52:02 Magi smbd[3352]: [2010/11/11 03:52:02,  0] lib/util_sock.c:539(read_fd_with_timeout)

Nov 11 03:52:02 Magi smbd[3352]: [2010/11/11 03:52:02,  0] lib/util_sock.c:1491(get_peer_addr_internal)

Nov 11 03:52:02 Magi smbd[3352]:  getpeername failed. Error was Transport endpoint is not connected

Nov 11 03:52:02 Magi smbd[3352]:  read_fd_with_timeout: client 0.0.0.0 read error = Connection reset by peer.

Nov 11 03:52:02 Magi smbd[3352]: [2010/11/11 03:52:02,  0] lib/util_sock.c:738(write_data)

Nov 11 03:52:02 Magi smbd[3352]: [2010/11/11 03:52:02,  0] lib/util_sock.c:1491(get_peer_addr_internal)

Nov 11 03:52:02 Magi smbd[3352]:  getpeername failed. Error was Transport endpoint is not connected

Nov 11 03:52:02 Magi smbd[3352]:  write_data: write failure in writing to client 0.0.0.0. Error Broken pipe

Nov 11 03:52:02 Magi smbd[3352]: [2010/11/11 03:52:02,  0] smbd/process.c:62(srv_send_smb)

Nov 11 03:52:02 Magi smbd[3352]:  Error writing 75 bytes to client. -1. (Transport endpoint is not connected)

Nov 11 03:52:02 Magi smbd[3352]: [2010/11/11 03:52:02,  1] smbd/service.c:1226(close_cnum)

Nov 11 03:52:02 Magi smbd[3352]:  wes-pc (192.168.11.117) closed connection to service Testshare

 

When I edited the /etc/samba/smb.conf and added:

smb ports = 139

under the [Global] section it seems to have resolved all the slow browsing and timeout issues but it still causes the daemon to exit. I'll post as I find more.

 

No Idea, tried the samba ports, also tried 5.0 beta2 to see if it was a bug in samba, disabled client NIC power management, changed the network security policies on Windows 7 Ultimate x64, and even tried an Intel Pro100S adapter. At wits end with this as all my clients are Windows 7.

 

Nothing worked, here's my syslog

syslog.txt

This is a shot in the dark, but try disabling IPv6 on your Win7 box.

This is a shot in the dark, but try disabling IPv6 on your Win7 box.

 

No luck

Do you have a fixed IP address for you unRAID server?  I have a windows 7 machine that would not see the unRAID server sometimes.  After I changed my unRAID server to a fixed IP address, my windows 7 machines always sees the unRAID server.

 

 

CJ

Do you have a fixed IP address for you unRAID server?  I have a windows 7 machine that would not see the unRAID server sometimes.  After I changed my unRAID server to a fixed IP address, my windows 7 machines always sees the unRAID server.

 

Tried it with no luck.

This is getting weirder and weirder.

 

I gave up on unraid for a bit and switched over to some other NAS products and was going to reload WHS. One of the products is giving the exact same issues but happened to notice that for a brief second the workstation looses network connectivity. Thinking it might be the Buffalo WZR-HP-G300NH (with DD-WRT) I tested with another cat5e cable connected directly to a gigabit switch bypassing the router the same issues occured. I'm pretty much convinced I have some wierd gremlin inside of this PC which only exists on large file copies.

 

Fired up another Windows 7 Home workstation and the 3rd party NAS works fine now. Back to reloading unraid and testing more.

After a full parity rebuild it still fails on a different workstation. Reinstalled WHS on the same hardware (well besides a 1TB EADS boot drive) and no issues at all copying the data back. I have shelfed using unraid for now as it's been two weeks down my NAS

Try going into Programs and Features > Turn Windows features on or off (left hand side) > choose to remove Remote Differential Compression and reboot.

Try going into Programs and Features > Turn Windows features on or off (left hand side) > choose to remove Remote Differential Compression and reboot.

 

Gave that a whirl with no luck, and tried setting the desktop nic's card speed, but I finally tracked down the main workstations issue. The realtek driver was recently upgrade a few weeks ago to one pushed out via Windows Update. The driver was issued by Realtek and not the 2008 driver included with Windows 7. After rolling back the network card driver (did a fresh install of windows to test with), the network traffic was nice and smooth instead of 10second pulses to 100% traffic which was causing timeouts.

 

Of course I'd find this after migrating 4TB of data back to WHS. I'll grab some smaller disks and setup another unraid install to test to see what issues remain.

Thanks everyone for all the tips and suggestions.  I have tried everything listed here and have not had any luck whatsoever.  I believe for me the issue is on the unraid server and not with the network card.  There is a possiblity it is a networking issue.

 

I can transfer files to one disk in the unraid.  Then when I go to transfer to a different disk, it times out.  Using terra copy the error message reads " The specified network name is no longer available."  This happens directly after using the unraid.  The only thing i can see is that the new drive I am trying to transfer to is asleep. 

 

Here is a guess.....  the drive is asleep so the unraid cannot start accepting the data until it spins up and initializes.  But by the time that happens my network card times out. 

 

As soon as I get this error - I can repeat trying to move the exact same file to the exact same location - and it works like a champ......

 

So how can i adjust the time out time?  Is that possible?

I have a good idea as to why this is happening as I have the same annoying problem. I think its due to a lack of space on the server.

 

I have 12.7TB on my server with only 2.1TB free (according to Windows 7). The first attempt at a transfer fails but I noticed that one the hdds activity light on the server is on now continuously on for a very long time. Checking the main unraid page shows that the hdd is continuously reading. Maybe its trying to calculate the space required for the file?? When it finally stops I try the transfer again and it works. "discovering items" is a bit slow but it works.

 

I then deleted some files until there was about 3.1TB of free space on the server. I then did a transfer and it works perfectly first time!!!! The file was "dicovered" by Windows instantaneously. Until the free space reached 2.54TB and then back to this annoying problem.

 

One way around this was to set the allocation method to "fill up". The "discovering items" is still a bit slow but it works.

 

Another way would be to add another hdd which I haven't tried yet.

Pras, What is your min space set to? It should be at least twice the largest file you will ever write.

I have set it to 80000000 but this doesn't have any effect on the problem. I only recently set this limit. Maybe the system needs a reboot? But I definitely think its an overall free space problem,

I have set it to 80000000 but this doesn't have any effect on the problem. I only recently set this limit. Maybe the system needs a reboot? But I definitely think its an overall free space problem,

I do not know when the min-free is read.  It might require a stop and re-start of the array.  It might only be read when the array is first booted, or it might take effect immediately once saved.

 

Joe L.

I set the min free space and it sorted out the issue with the "fill up" problem that I mentioned in another thread. Unfortunately I can't reboot yet as I have installed my parity drive after I put all of the data on the server. I will check after a reboot. But surely the server software should be keeping a close eye on all of the hdds free space without the need to carry out a massive pre-read before each file transfer?

Finally found a thread describing the exact issue that I've been having transferring files to my UnRAID server from Windows Vista x64.  I have 6x 2TB drives in my array and 1x 1TB drive that is empty.  Static IPs on both machines.  Basically, it's exactly as the others have described and I find the free space issue to be intriguing as I could never exactly pinpoint the problem.  One thing is for sure, my drives are spun down 99% of the time that I have a problem and either browsing to the disk in question (I don't use disk shares) or trying a second time solves the issue.  Once I've made a recent transfer, remaining transfers work with no issue.  If a drive is spun down and I want to transfer files to it then I also notice very slow transfer speeds at the beginning of the copy as well as Windows not correctly displaying the transfer rate or estimate time to completion.  I also did some checking on various Windows forums and others have described similar issues when copying large files to network shares in Windows OS's post XP.

 

Anyway, I've also got a Realtek RTL8168C(P)/8111C(P) Family PCI-E GBE NIC with a 2009 driver from Realtek installed (3/17/2009, 6.218.317.2009) so I'm going to try rolling back the driver to a Windows version and see if that helps.  I'll report back later once my server is running and I can do the test.

I have rebooted and tried a to rip a blu ray to the unraid and the same problem occurs.

 

I will preclear some hdds and see if I can resolve the problem by adding more space. It will take about 25 hours to clear the hdds!

Well, so much for rolling back the driver for my NICs.  I couldn't get the Windows driver to work...not sure why.  Anyway, I updated to a newer version from Realtek released on 12/22 and so far no issues.  I'll keep reporting back though.