Server Fault is a question and answer site for system and network administrators. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

I am building a new network for a company from scratch (50 user/Computer)

They need

1-Domain Controler ( Only one Domain at the moment )

2-MS Exechange Server

3-File Server

4-SQL/Database Server "Not Too Much load"

I am planning the following:

Two IBM Servers and every Server Has(2x4core server-64 GB Ram-300GB Raid1 Hard Disk)

SAN System with Raid 5

Using Hyper-V and Fail over cluster (2 nodes) Can i use the following

First Server ( DC, File Server), Second server (Secondery DC ,Exchange, SQL)

is that possible ? is that advisable ? and Why ?

share|improve this question
I'm sorry, but why don't you just use SBS? – JohnThePro Feb 13 '12 at 18:04

Possible, Yes, Advisable, No.

The setup you describe would work, but if you think about disaster recovery and plan for failures you will see the deficiencies. You need to have redundancy and have a good backup plan (offsite!). With the configuration you describe I don't see any benefit from clustering, also, it will not work with both domain controllers virtualized (see comment from TomTom below).

For most business cases I would agree that MS SBS is a good option. It really depends on what the needs of the business are, but I can't see any business that isn't concerned about continuity of operations.

share|improve this answer
Yup - for 50 users, get SBS on one server, and make the other server a DC as well. – mfinni Feb 13 '12 at 18:25
Acutally no, it is not possible to virtualize domain controllers and do clusering. YOu can not start a cluster without domain controllers up. – TomTom Feb 13 '12 at 18:53
@TomTom - yes, I ignored the clustering part because it doesn't make any sense to me in this configuration. – jhayes Feb 13 '12 at 18:59
As I say in my answer. But it still is in his configuration. And it is a bad bad trap. – TomTom Feb 13 '12 at 19:08

It is stupid.

Sorry, that blunt.

See, cluster service NEEDS dc.

What happens now:

  • Power fails.
  • Power comes back.
  • Cluster does not start any VM's because there is no DC to pull the configuartion from.

End of game - have fun handfixing that.

every Server Has(2x4core server-64 GB Ram-300GB Raid1 Hard Disk)

Dead. This is a slow disc system. I run one like that (2x4 core) and ahve 8 discs in a raid 10 and I am still too slow. And that is for a dozen users. Plus another 10 discs for a database server, but ok - I deal with hundreds of gigabytes.

My advice:

  • Make the phyical machiens DC's.

This is waht I do in my office (2x 1hexacore with 16gb RAM). Do NOT virtualize the domain controllers - at least keep 2 physical. Otherwise you are in a world of pain when something blows. I also keep file server physical (windows core function). DFS, DNS, DHVP are done physical. has a good reference to the pain you get into.

share|improve this answer
Agreed, keep some DCs physical, as well as any core networking systems. – Tatas Feb 13 '12 at 17:17
Thanks For that, Got it now, actually this will be my first "Fail over cluster" implementation... Anyhow, About virtualization it seems nobody advice me to virtualize any server No DC, No File Server, So What To virtualize ??? Does SBS 2011 has some Limitation??? – ssaabbaa Feb 15 '12 at 7:42
I vuirtualize everything except .- i keep dc's physical, at least those running the infrastructure, file servers aetc., areo nthe physical box, too (only overhead with virtualization on same box). FInancial trading aopps pyhsical. OTherwise I love virtualization. I think I have 50 or so vm's i my infrastructure. – TomTom Feb 15 '12 at 9:51

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.