Clustering Servers Over a WAN

Hi All

my setup is a little bit confussing but I am going to do the best i can. We currently have two Datacenters, lets call them DC1 and DC2. Both sites have a FC SAN and Multiple servers running ESX 4.1.

What i would like to do is create a MS 2008 failover cluster for DHCP services and File print services.

But the issue is that in order get the Cluster working it needs Shared storage and because I am using Vmware and I have two different sites the cluster wont function properly.

So my question is can MS 2008 Clustering and Vmwre function over a WAN?
OTNAdminAsked:
Who is Participating?

[Webinar] Streamline your web hosting managementRegister Today

x
 
Danny McDanielConnect With a Mentor Clinical Systems AnalystCommented:
It sounds like your best bet would be to look into some kind of replication between the two DC/SAN's.  Veeam's backup solution does replication, too, but you have to manually failover to the replicated system or maybe setup some kind of monitoring that triggers a powershell script to start the replicated instance up.
0
 
khuphucCommented:
in this condition of FC San on each site, you can't dedicate physically the quorum disk, then you can't do it

maybe in iscsi configuration you can reach it, but it been so hard to configurate

if you want to create a high availability server use the vmware services, maybe better
HA;
Fault Tolerance
etc

regards
0
 
Luciano PatrãoICT Senior Infraestructure  Engineer  Commented:
Hi

How many ESX servers do you have? If you have more than one ESX server, and have vCenter you do not need to install MS Clusters. This is a waste of MS licenses and money. You can do this Clustering and High Availability with vCenter(creating some ESX Cluster with HA and DRS).

And if you need a other site to replicate your Environment, then like danm66 suggests, Veeam is a very good product(and cheap) that can create your Replications between sites.

Jail
0
Problems using Powershell and Active Directory?

Managing Active Directory does not always have to be complicated.  If you are spending more time trying instead of doing, then it's time to look at something else. For nearly 20 years, AD admins around the world have used one tool for day-to-day AD management: Hyena. Discover why

 
kevinhsiehConnect With a Mentor Commented:
You can use Windows clustering in a configuration that is known as a stretch cluster. Yes, it can be done, but I would consider it a specialized use case. It requires a way to replicate the storage.

http://downloads5.eventpoint.com/teched2009/slides_europe/SVR319_Christensen.pptx

There are other options such as Marathon Technologies or Double-Take.

My thought is that DHCP doesn't need to be clustered. Just use multiple DHCP servers and split up your scopes. You can configure your routers to forward DHCP requests to another datacenter if you need to.
http://technet.microsoft.com/en-us/library/dd296651(WS.10).aspx

For the file server, a standard Windows stretch cluster is a big deal. Perhaps an easier solution is to use DFS replication or something like Double-Take to keep your file servers in sync. Failover can be manual or perhaps automatic, and it doesn't take more than what you have right now.

For the print server, Double-Take, Marathon, Windows Clustering, keeping a copy of the VM at DC2 are all options.
0
 
VirtaliciousCommented:
Look.... I know you have your heart set on doing this... but please let me tell you that it is not the most efficient use of resources.

You should reconsider just allowing DRS/HA or even FT for this situation as they would allow a much more robust environment with many less restrictions.

Just trying to be the voice of reason...

-Virt
0
 
OTNAdminAuthor Commented:
Thank you all for your coments, lets go over them.

Virtalicious:
DRS/HA would work fine in ONE Datacenter but not two. The requirements for DRS/HA and even FT is Shared storage and because we are using FC, I can not extend My SAN to both Datacenters.

kevinhsieh:
You mentioned Doubletake, we currently have doulbetake in are environment. let me ask you this could we create Two MS Fail over cluster one at one DC and One at another, then using a double take agent replicate the shared disk from one DC to another?
0
 
Gerald ConnollyCommented:
Microsoft cluster used to require that both nodes were in the same subnet! (and that there was a latency limit) which normally makes it difficult to run the cluster across sites - its not that its impossible just that its difficult.

Split site clusters always cause worries about the split brain scenario!
0
All Courses

From novice to tech pro — start learning today.