Want to protect your cyber security and still get fast solutions? Ask a secure question today.Go Premium

x
  • Status: Solved
  • Priority: Medium
  • Security: Public
  • Views: 686
  • Last Modified:

How could changing the Binding Order affect server operation

How could swapping the binding order of the Remote Access WAN Wrapper and the Internal NIC adapters under the WINS (TCP/IP) protocol in the Remote Access Server Service affect the operation of the Server?

Currently the internal NIC is below the Remote Access WAN Wrapper.

Background Info:
Our Exchange 5.5 (SP4) server is a WinNT 4.0 (w/ SP6a) BDC server and also our RAS server.  Multiple remote sites connect to both a Windows 2000 PDC and this WinNT 4.0 BDC server at our location (via LAN & WAN).  Each remote office has a WinNT 4.0 BDC and LAN and our entire organization is under a single domain.  There are also multiple mobile users that frequently connect to the network via dial up to this WinNT 4.0 RAS server.  DHCP on our Windows 2000 PDC hands out IP addresses to these remote users.  These RAS addresses are not excluded from the general pool of IP addresses given to the users on the local LAN.

Recently the Exchange MTA service failed when it attempted to re-start and in turn caused the dependent IMC service to also fail.  Re-booting the server allowed all Exchange services to restart properly.

When the MTA service failed to re-start the Event Viewer logged several errors.  Errors 2219, 2206, and 2207 suggested the MTA service had not shut down cleanly – however there were no other indications that it did not shut down properly.  Errors 9321, 9266 and 9406 pointed me to KB article 279415 (we did not recently install Exchange 5.5 SP4, but might that still be relevant?) which recommended changing the binding order of all services so the internal NIC is at the top.  This server has one service (Remote Access Server service) that is not configured properly (as referred to in my question).  I need to be certain that “correcting” the order will not cause other issues to arise.

If changing the binding order of these adapters does not fix the “intermittent” MTA service startup problem, then there is an Exchange server 5.5 post-SP4 MTA Fix available that might help (KB article 283238).  We have very few of the MS security fixes, etc. installed on this server.

Thanks in advance for any assistance.
Kevin
0
kbrak
Asked:
kbrak
  • 3
  • 2
1 Solution
 
andyalderSaggar makers bottom knockerCommented:
It's to do with RPCs only binding properly to the first listed adapter and Exchange uses local RPCs to talk to itself.

Correcting the binding order shouldn't cause problems, I've had a couple of customers with this error and the binding order fix didn't cause them any problems.
0
 
kbrakAuthor Commented:
Thanks for the info.

One thing I neglected to mention which may be related to the way the binding order is currently set up is whenever we re-boot this server, we must unplug the network cable otherwise we get the following errors in the Event Viewer at startup:
Server Service error - "The server could not bind to the transport\Device\NetBT_CpqNF31"
Workstation Service error - "Machine name is not a valid computername"  "A duplicate name exists on the network".

Are these errors related to the way our binding order is set and if so, will making the proposed binding order change possibly correct these errors too?

Thanks in advance for any assistance.
Kevin
0
 
andyalderSaggar makers bottom knockerCommented:
I would say it's far more fundamental than the binding order if you have to unplug the network cable to get it to boot. Are you sure you only have one network interface? - just that the netflex3 is often teamed and then the order is critical as someone could have bound something to an individual card.

They won't give you the hotfix unless you can get it to boot cleanly with exchange services disabled.
0
What does it mean to be "Always On"?

Is your cloud always on? With an Always On cloud you won't have to worry about downtime for maintenance or software application code updates, ensuring that your bottom line isn't affected.

 
kbrakAuthor Commented:
I verified that there is only one nic in the server.  The only adapters are the nic (Integrated Netelligent 10/100 TX PCI Bus 1) and a modem (Digi Acceleport RAS 4/8 Adapter).

After re-booting this server with the network cable unplugged, we are then able to plug in the cable and everything works fine.  We only get the errors if the cable is connected at boot.

The only protocol installed is TCP/IP.  In the bindings it shows protocols TCP/IP, WINS Client(TCP/IP) and Remote Access WAN Wrapper bound to the apadters as shown below:

(3)Remote Access WAN Wrapper
     Remote Access Wan Wrapper

(5)Remote Access WAN Wrapper
     Remote Access WAN Wrapper
     TCP/IP
     WINS Client (TCP/IP)
(1)Integrated Netelligent 10/100 TX PCI Bus1
     TCP/IP
     WINS Client (TCP/IP)
------------------------------------
Bindings displayed by All Services:

NetBIOS Interface
     WINS Client (TCP/IP)
         (1)Integrated Netelligent 10/100 TX PCI Bus1
         (5)Remote Access WAN Wrapper
Remote Access Server Service
     WINS Client (TCP/IP)
         (5)Remote Access WAN Wrapper        
         (1)Integrated Netelligent 10/100 TX PCI Bus1
Server
     WINS Client (TCP/IP)
         (1)Integrated Netelligent 10/100 TX PCI Bus1
Workstation
     WINS Client (TCP/IP)
         (1)Integrated Netelligent 10/100 TX PCI Bus1

The adapter order below the Remote Access Server Service is the one that appears to be incorrect.  I don't have a good understanding of how the bindings, adapters and protocols affect the operation of the server.

I was asked to research how switching the order of the adapters will affect server operation and to basically "state my case" before any changes would be allowed.  I am struggling to come up with any definitive answers.  You mentioned having made this change yourself without experiencing any problems.  Any suggestions on what I might look for to justify this change and to re-assure everyone that it won't crash the server.

I appreciate your assistance very much.
Kevin
0
 
andyalderSaggar makers bottom knockerCommented:
0
 
CleanupPingCommented:
kbrak:
This old question needs to be finalized -- accept an answer, split points, or get a refund.  For information on your options, please click here-> http:/help/closing.jsp#1 
EXPERTS:
Post your closing recommendations!  No comment means you don't care.
0

Featured Post

VIDEO: THE CONCERTO CLOUD FOR HEALTHCARE

Modern healthcare requires a modern cloud. View this brief video to understand how the Concerto Cloud for Healthcare can help your organization.

  • 3
  • 2
Tackle projects and never again get stuck behind a technical roadblock.
Join Now