Home / server / Questions / 200763
Accepted
Hugo Garcia
Asked: 2010-11-12 05:11:44 +0800 CST

Why does the W2k3 TCP/IP stack gets corrupted?

  • 772

I have a W2K3 server that every time an update restart the server, at boot the tcp/ip stack gets corrupted.

This makes the server unable to do any kind of networking, and when i issue an 

 ipconfig

on a Command Window, the result is empty. to solve this problem i have to issue the following command and restart the server:

 netsh int ip reset <logfile>

This is causing a full network outage, because almost all services depends on the Domain Controller. when the server restarts i have to configure again the tcp/ip settings.

Another question regarding this issue is why the clients and Applications do not use the secondary domain controller, that is suposed to take over when the primary one fails?.

here are the FSMO roles assignations, SSCZDC01 is the server that fails>

dsquery server -hasfsmo rid
"CN=SSCZDC02,CN=Servers,CN=Default-First-Site-Name,CN=Sites,CN=Configuration,DC=
XXXXXXXXXX,DC=com"  

dsquery server -hasfsmo pdc
"CN=SSCZDC01,CN=Servers,CN=Default-First-Site-Name,CN=Sites,CN=Configuration,DC=
XXXXXXXXXX,DC=com"

dsquery server -hasfsmo name
"CN=SSCZDC01,CN=Servers,CN=Default-First-Site-Name,CN=Sites,CN=Configuration,DC=
XXXXXXXXXX,DC=com"

dsquery server -hasfsmo schema
"CN=SSCZDC01,CN=Servers,CN=Default-First-Site-Name,CN=Sites,CN=Configuration,DC=
XXXXXXXXXX,DC=com"  

dsquery server -hasfsmo infr  
"CN=SSCZDC02,CN=Servers,CN=Default-First-Site-Name,CN=Sites,CN=Configuration,DC=
enserbic,DC=com"

any insights or comments would be very appreciated, because i'm lost regarding the possible origin of this problem.

thanks very much for your time

windows-server-2003 active-directory tcpip corruption

2 Answers

  1. Best Answer

    Have you done a recent virus scan with an up to date virus scanner on the problem server? Is the server up to date on it's windows updates?

    I'm guessing from your tags that your domain is a Windows 2003 active directory domain. I also take it that you have a single domain forest with two DCs.

    It sounds like this server is the owner of all the FSMO roles on your domain (each of the five FSMO roles can only be on one server at a time).

    For info about FSMO roles: http://technet.microsoft.com/en-us/library/cc773108(WS.10).aspx

    You can determine if your problem server holds the FSMO roles like so on the command prompt:

    dsquery server -hasfsmo rid
dsquery server -hasfsmo pdc
dsquery server -hasfsmo name
dsquery server -hasfsmo schema
dsquery server -hasfsmo infr

    If your problem server holds all the roles above, I would transfer the roles to your good DC, as per: http://technet.microsoft.com/en-us/library/cc781578(WS.10).aspx

    This will probably solve the problems you get when the problem server drops out. Once you have transferred the roles, you can take the problem DC offline and do some diags on it, or as a last resort rebuild it.

    • 3

  2. Take a look at the network drivers and see if there are updated ones. It could be a bug in them that triggers it.

    Do you see anything in the event logs during the restart? there probably isn't especially if it occurs before the logging service gets started but you never know.

    We had something similar to a couple of laptops once where it completely lost the network and it would occur (eventually like a month later) even after a reinstall. Finally figured out since it was a group of them with sequential serial numbers that it was bad components that were somehow screwing with the OS.

    • 1