Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[michaeljc70]

I get this message when copying large files sometimes.  I don't get it on smaller files.  By large files I mean over 500MB.   Nothing else has changed in my setup. I am using a wired network.  I have the latest firmware (1.03).  The drives are still connected (I can browse them from the same machine I am trying to copy from), so they aren't really disconnected.

I tried:

-mapping drives to the volumes
-restarting the nas
-copying various files from various locations
-disabled virus software
-verified sufficient HD space
-Tried using windows explorer and robocopy to copy

Any ideas?

[kennyg]

I'm having the same problem.  I started noticing it with large files, but I think it is more of a timing issue.  All of my windows boxes and my mac lose connection to this device.  They all reconnect automatically, but sometimes this happens during a file write, which causes an error.    This is a pretty frustrating situation and unacceptable of a NAS.  I have been trying to find something wrong with the network, but so far everything else on the network is running smoothly for both wired and unwired.

When I claim it's a "timing issue" I mean that the device reports an error and if the copy program does not have error recovery, the process fails.  I first noticed it with a large DVD data file (NOT FOB).  When I installed TeraCopy, the process completed, but the verify process kept getting stock on the CRC verification, not failing, but getting stuck.

After that, I noticed all my machines lose drive mapping.  When you try to remap, windows reports it's already mapped (the problem disappears), then a few minutes later, it fails again....

I'm not sure what the next steps are, I hope that one of the dlink moderators respond, because otherwise I'm just going to have to call it a week and get a different NAS solution.

- Ken
- Ken

[kennyg]

Before anyone Asks:

Firmware: 1.03
Drives: Barracuda 1.5 T Raid 1
IP:  Static (Both Test Machine and DNS-321)
Drive Spin Down: Disabled
Jumbo Packets: Disabled
Switch:  1 Gig Cisco SG 100D-08

NAS Performance Checker using 1 Gig File
     Pass 1: 7.11 M/Sec
     An Error occurred:  The specified network name is no longer available
 
NAS Performance Checker using 100M File
     Pass 1: 7.61 M/Sec
     Pass 2: 7.73 M/Sec
     Pass 3: 7.67 M/Sec
     Pass 4: 7.73 M/Sec

 
BTW: This seems a bit slow to me, but the test box has only 100M NIC.


So, I copied

[kennyg]

For those interested.

My problem has been reduced considerably since I reconfigured Macafee on all the windows boxes to treat all machines on my local network as a "trusted network".  Apparently, either the Macafee security is broken in this regard (meaning either allow or disallow access across machines), or is sufficiently delaying connections so that they appear to be temporarily disconnected.

However, the problem is not completely eliminated and I still had large file problems as did an earlier poster.

I resolved those by using the LMHOSTS.sam file on all boxes (they were already all static ip) to resolve DNS names for all network shares.   This likely solves or reduces the problem because of the lack of DNS resolution for the shares.  If you don't want to do this, you can also use the ip address to map your shared drives.

Now, most of my problems are solved.  It is still disturbing that my mac gets temporarily disconnected from the NAS, which could be due to the wireless connection.  Still, I never lose wireless connection from the same wireless laptop to my apple time machine.

There are no ip conflicts or dhcp challenges on this network, and I have no other network performance problems.

After all the solutions were applied I have the following benchmarks to report writing to the DNS-321

Using 1g cisco switch
From Windows XP boxes with 1g NICS:   Average 14.50Meg for 1 gig files write, average 9.5 for read
From Windows XP Boxes with 100M Nic.  Average 7.15Meg for 1 gig files write, average 3.5 for read
From a write perspective this is much faster than network shares for write but about equivalent for read.

I'm curious that read percentages are so much lower, but I'm guessing there is a write through cache on the NAS that buffers some of the writes and that there is no read ahead caching on the DNS-321

Conclustion:  Something is going on with this device either in design or on my network that makes it perform differently (unavailable) than other network storage devices (XP Shares, apple time machine). 

Therfore, I am going to use it not as a reliable NAS, but only as a backup device using backup utilities with verify.

- Ken