Go Premium for a chance to win a PS4. Enter to Win

x
  • Status: Solved
  • Priority: Medium
  • Security: Public
  • Views: 346
  • Last Modified:

Replace Active Directory for DNS replacement

I'm looking for recommendations.

I'm currently running a Microsoft Active Directory based internal network with the MS servers providing DNS resolution for both internal and Internet names.  We use a split-DNS zone for the public domain name (e.g., "mycompany.com" , as well as a private DNS zone for AD (e.g., "mycompany.local")

I have 1 remote branch with a single Domain Controller - which is also the sole DNS entry for the local devices.  If that DC is offline, there is *no* DNS resolution for the branch, and "the Internet is down".

I can't add a public DNS to those users' configuration, because then they start having AD issues with the local/corporate network, as well as access to services like Exchange for email.

I think the best solution is a 2nd DC/DNS server at that location - it can be PC-class hardware for all I care; just something to add a 2nd DNS entry in case the primary DNS is unavailable.

Are there alternative configurations to this?

Get rid of Active Directory?  What are the alternatives in the non-MS realm?
0
snowdog_2112
Asked:
snowdog_2112
  • 3
  • 2
  • 2
  • +2
2 Solutions
 
Seth SimmonsSr. Systems AdministratorCommented:
it's always a best practice to have redundant domain controllers in each site for situations like this

on the one hand, adding a cheap box with linux configured with bind to pull dns from the domain controller solves most of the dns part, but you will still experience logon failures and/or delays because of the domain controller being down at that site

better off with a second domain controller for complete redundancy
0
 
Brad HeldCommented:
I agree that a second DC would be best in this site, particularly if wan traffic to the central office is undesirable.

Option 1: add a secondary DC/DNS server at the site, configure the clients to use the secondary DC as secondary DNS

Option 2: Add a dc from another site as secondary dns for clients

One of the issues with using say Bind is the stickiness for the clients to hang on to the secondary once the secondary responds which means the client won't automatically fail back to the primary
0
 
arnoldCommented:
Adding to the above comments deals with does the remote location use the DHCP on the DC or does the router/fireweall allocate IPs?
Presumably there is a VPN between the two locations.

One option is to publish via DHCP assignment both the local and the remote DNS server.
the difficulty deals with the DNS queries will be "load balanced" between the local and the remote DNS.

This way if you loose the local DC on the remote site, the logins/etc. will be slower because the requests have to be sent through the VPN to the remaining DC.

Depending on your AD DC setup, you could require the remote branch sistems to always query the local DC for data before sending requests to the HQ DC.
0
 [eBook] Windows Nano Server

Download this FREE eBook and learn all you need to get started with Windows Nano Server, including deployment options, remote management
and troubleshooting tips and tricks

 
Jeremy WeisingerSenior Network Consultant / EngineerCommented:
@Arnold "the difficulty deals with the DNS queries will be "load balanced" between the local and the remote DNS."

Maybe I'm misunderstanding you but the DNS client in Windows does not load balance queries. It will only use the secondary if the primary is unreachable. And it will only switch back to the primary (or continue down the list of DNS servers) if the secondary is unreachable or at DHCP refresh, NIC status change, or a reboot.
0
 
arnoldCommented:
Prior experts comments addressed the preferred route of having redundancy via two DCs at each location.

My addition was to deal with the situation as is where there is only one system available as a DC at the remote location and how to mitigate the loss of said system while maintaining the functionality of the remote systems. login, access to the internet, etc.

With the current setup which is all or none, the bandwidth cost of DNS queries going through the VPN will still provide some reasonable fall back position.

Possibly using/setting up a virtual VM DC (which must never be restored from snapshot/backup) on one of the "workstations" at the remote office could provide "additional" fault tolerance.
0
 
snowdog_2112Author Commented:
Great comments folks - thank you so much!

What would you all recommend if we were to GET RID of Active Directory at that location?

Assuming we are open to removing AD at the branch, and still have AD at the datacenter (yes, there is a tunnel between locations) for services like Exchange.

In terms of options, then:

1. Keep AD, and add redundancy to DNS at the branch (a 2nd DC or simple DNS with zone replication to the DC)

2. Get rid of AD and allow zone transfers from the datacenter DNS servers for internal DNS zones.

3.  Something else altogether?  Something completely non-Microsoft?

We can configure DNS zone transfers from the datacenter, but the question then becomes user authentication.  There aren't really any local "shares" at that location - mostly PC's accessing resources at the Datacenter or Internet.
0
 
Seth SimmonsSr. Systems AdministratorCommented:
you still need domain controller(s) to authenticate, processing of group policies, etc.
having none there - as i stated earlier - will cause problems when accessing resources
adding an additional domain controller at that location is the simplest way to go.  if that tunnel between locations goes down, then you have a problem
0
 
snowdog_2112Author Commented:
That is my feeling as well.

I am looking to justify the current situation (with an added AD/DNS server) essentially by proving every *other* solution is not viable, if that makes sense.

"We could do 'X', but these are the disadvantages.  We could do 'Y' but then we have these other disadvantages".

Thanks!
0
 
snowdog_2112Author Commented:
There certainly is something to be said for Windows reluctance to effectively use the two DNS servers in its own configuration (whether round-robin, failover/failback, or user-definable) is a concern.

Given that, it seems a second DC is the best solution - even with the added cost of the OS.  That cost will be saved in administration and productivity over time.

Thanks for all the thoughtful input!!
0

Featured Post

How to Use the Help Bell

Need to boost the visibility of your question for solutions? Use the Experts Exchange Help Bell to confirm priority levels and contact subject-matter experts for question attention.  Check out this how-to article for more information.

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