Link to home
Start Free TrialLog in
Avatar of gee_gdavid
gee_gdavid

asked on

Error 53. Computer Name not found using a DOS 6.22 network boot client

Using network monitor I see the NS request from the client. I also see the reply from the server. I queried the MAC address to verify the NICS   I see in the capture.

The weird part is that I used this and other similar floppy disks and it worked in the past. I have turned on auditing. I have run MS Baseline Security Analyzer.

Thoughts, suggestions, help?
Avatar of CYBERUG
CYBERUG

I often get this error when having to do a ris install across a cisco router, to address the problem I normaly reduce the 10/100 nic down to 10mb so as to pass smoothly through the cisco hub or router.
If you have a config disk for the nic yous it to reduce to 10mb and reconfigureafter youve completed your install. (or leave at 10mb)(?).
Where do you get this problem? If it's when mapping a drive, just be sure the name of the server and the name of the share do not exceed 13 characters and contain no special characters or spaces. Also, you can't map drives with ip instead of name in dos, like \\192.168.0.x\sharename.

Give more info please. I often use network boot floppies to ghost computers to and from servers, for example, but there are many thing to consider, like the server o/s, the kind of link you want, the nic, etc.

If you have the same boot disk that was working before and it uses tcpip, do you try to map drives on a xp or 2003 computer? Then you would have to adjust some things, like disabling SMB signing in windows.

Luc
Avatar of gee_gdavid

ASKER

Here is some additional background.

The UNC is unchanged. For the last six months this UNC (server and share name) have not changed. The bootable floppies have not changed. The network infrastructure has not changed (no routers, using a switch). The change that was introduced was Windows update and the Microsoft Baseline Security Analyzer.

As I mentioned in the initial post, the client sends an NS request. The server responds to the request. The client sends the request one time only. The next traffic from the client is an ARP request to the gateway. At that point all traffic ceases and the client posts the Error 53.  The same client floppy can connect to a share on a Windows 2000 server.

I am suspicious of the anonymous settings on the 2003 server which I reversed to the high risk setting. The server was rebooted but the change was unproductive.

More ideas?
The SMB signing is required by default on 2003 server. It is enabled but not required on 2000 server. It is not supported on DOS or win9x.

Check:

HKLM\System\CurrentControlSet\Services\LanmanServer\Parameters\

REG_DWORD: RequireSecuritySignature=0 (default=1)

Luc
Sadly, the SMB parameter is set to 0.

HKLM\System\CurrentControlSet\Services\LanmanServer\Parameters\

REG_DWORD: RequireSecuritySignature=0 (default=1)


Thus, the problem persists. More ideas?
Try to force "enable netbios" in tcpip properties (under wins tab).

Luc
Here is an update. I looked more closely at the network capture and found, in the NS reply that it includes two references. Remember that the source computer (Win 2003 Server) has two adapters on separate networks. 192.168.0.x and 192.168.100.x. The subnet is 255.255.255.0.  The order of the name references lists the "remote" network first. This is why I see traffic pointing to the gateway. Hmmm.  Two questions:
Why would the client work for months and now fail? Why would the server offer a single reply with two resolutions?

Ooops, a more important question: How do I fix this?

DW
ASKER CERTIFIED SOLUTION
Avatar of lstonge
lstonge

Link to home
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
Start Free Trial
BTW ms-dos is not mentionned in this q826157 but it applies to it also.

Luc