I am posting this here due to the fact that the base system will be Server 2008.
I am redesigning a network and as such am exploring the current system against the desired result.
My colleague and I have found a point of disagreement and so I would very much appreciate everyone's thoughts...
We have currently 11-13 servers with individual roles. The Domain Controllers (2) only concern themselves with that role (DHCP, DNS, log in, etc), while we have a server for file shares, one server for Exchange, one as the Proxy Server, one for printers, etc.
The disagreement is as follows:
Should we consolidate into a small body of servers (3 or 4) that handle all rolls and thus reduce confusion of which server should respond to a call? However we open ourselves up to potential network failure if a simple service fails on a server and it takes the server out (so, the Print Service collapses and removes the DHCP, DNS and login elements as the server reboots).
Or should we keep with a diverse group of servers, thus spreading the load across many units but then create potential confusion between servers; with each one being polled to see which one should respond?
I suppose in line with this is: do we include ALL servers into managing AD functions and if not then how do we select the ones that do and the ones that stay outside?
I am redesigning a network and as such am exploring the current system against the desired result.
My colleague and I have found a point of disagreement and so I would very much appreciate everyone's thoughts...
We have currently 11-13 servers with individual roles. The Domain Controllers (2) only concern themselves with that role (DHCP, DNS, log in, etc), while we have a server for file shares, one server for Exchange, one as the Proxy Server, one for printers, etc.
The disagreement is as follows:
Should we consolidate into a small body of servers (3 or 4) that handle all rolls and thus reduce confusion of which server should respond to a call? However we open ourselves up to potential network failure if a simple service fails on a server and it takes the server out (so, the Print Service collapses and removes the DHCP, DNS and login elements as the server reboots).
Or should we keep with a diverse group of servers, thus spreading the load across many units but then create potential confusion between servers; with each one being polled to see which one should respond?
I suppose in line with this is: do we include ALL servers into managing AD functions and if not then how do we select the ones that do and the ones that stay outside?