Domain/DNS synchronisation

[Taggs]

Here is my scenario. We have a web site on server ‘A’ that people connect to via a DNS entry. Server A’s web site connects to our mainframe via its MAC address through NDIS.

I have set up another server (server ‘B’) exactly the same as server ‘A’. If server ‘A’ fails all I need to do is change the DNS entry to point to server ‘B’ and people can carry on working. My problem is when I change the DNS entry on a domain controller, it takes around 5-10 mins for the four domain controllers to sync.

The DNS TTL is set to 0.

Is there any way to force a windows 2000 domain controller to synchronise its DNS. Or a way of changing all four domain controllers at once? Or a better way to perform this action?

Thanks

Taggs

[cluberti]

You could run "ipconfig /flushdns" followed by "ipconfig /registerdns"

[Taggs]

Does this sync the domain controllers?

I know you can do this on the PC's (250+ PC's is a lot of people to talk through and I want it to look seemless for the user) but if you are not connected to the domain controller that I updated then you have to wait the 5-10mins for the update.

B)

[cluberti]

DNS between AD domain controllers syncs on AD synch, unless configured otherwise.

[Taggs]

The problems I have is that it takes 5-10mins for the domain controllers to automatically sync

I want to be able to make the DNS change and then force/speed-up the update across the controllers so the users are not without the service for any length of time.

I not sure if I am making sense? Please let me know and I will try and explain a little better

Taggs

Edited July 23, 2007 by Taggs

[nmX.Memnoch]

I think the better question is: Is server 'A' failing that often? And if so, why?

[fizban2]

if i Server A is something that is important that it can't be down. wouldn't created some sort of web server farm or clustering the 2 machines so that if Server A was to go down taht Server B would take over and no configuration would need to be done on your part?

[nmX.Memnoch]

I'm in agreement on that suggestion as well. Use Network Load Balancing (assuming the web servers are Windows/IIS).

But, my question about Server A still stands. Is it really crashing that often or is this something you found out during testing?

[Taggs]

Thanks for the comments but comments like that are not really helpful at all for getting an answer to my problem.

if i Server A is something that is important that it can't be down. wouldn't created some sort of web server farm or clustering the 2 machines so that if Server A was to go down that Server B would take over and no configuration would need to be done on your part?

I have already gone through these solutions and they do not fit this problem. I also know what the problem is (network problem which does not happen very often and of which we have no control and I do not want to go off on a tangent talking about it).

Network load balancing will not work either as IIS site will still be running but mainframe connection dropped, therefore connections will still be put through to a server not working.

I want to provide the best service I can to my users and if I can sort this problem, the service will be improved.

Thanks

Taggs

Edited July 25, 2007 by Taggs

[DennisT]

Active Directory Sites & Services -> Sites -> Default-First-Site-Name -> Servers -> Server you modified -> NTDS Settings -> Right click on each object in the right pane and select "Replicate Now".

OR

You can install Specops Gpupdate and update everything from AD Users and Computers.