DHCP Scope Recommendations For Windows Server 2003

Are there any guidelines to follow when creating a DHCP scope in Server 2003?  For example, is it recommended to have 25%, 50%, etc. more IP addreses in your defined scope than you actually have nodes that need addresses leased?
Linds462Asked:
Who is Participating?
 
zefiroCommented:
My rule of thumb is to include as many addresses as you expect to need during the lifetime of the server, then add 10% so you don't get events in your event log.

 I also like to start and end the scope with a nice round number, so if I have to add a device that needs a static IP to the network, it is easier for my CRT-radiation addled brain to remember where not to put it.

Also, set-up scope options (Subnet, Gateway, DNS, WINS, etc) within the scope instead of using the global scope options, just in case you ever have to set-up a new DHCP scope, can save you a bit of work.
0
 
JamesDSCommented:
Linds462
Established practices in DHCP design dictate that the scope should consist of ALL possible addresses in the subnet.
You then use multiple Exclusion ranges on the scope to limit the number of addresses where each exclusion consists of say a 100 address block.

It is done this way because DHCP Scopes cannot be modified once created, but exclusions can be removed at will and each exclusion removal will extend the DHCP scope range by X addresses that are immediatly available.

As an additional precaution I also also split the DHCP Ranges across 2 servers and run both live at the same time for resilience - I have deployed this system successfully to several client, one with over 400 IP subnets - all served by 2 DHCP servers

Cheers

JamesDS
0
 
zefiroCommented:
Just wanted to note that JamesDS's answer is a good one.  While my answer focused more on reducing administrative hassles, James's answer would probably be considered best practice.  My customer base is pretty large and generally not technically advanced enough to even know what DHCP is, so my rules-of-thumb is designed to cause my clients and myself the least amount of headache over time while maintaining stability, security, etc.

Hope this helps
0
 
JamesDSCommented:
zefiro

Thank you for the comments. This is a case of large versus small clients. The large clients will want a best practice solution - the smaller clients will want something easy to administer. Either answer is correct, but yours would appear to be more suitable to the situation and therefore more deserving of the points.
Cheers

JamesDS
0
 
Linds462Author Commented:
Thanks guys.  I think we are going to go with a combo.  We are medium sized business, but are merging locations, so we wanted to make sure that we were going to have enough space since we are putting in a new system.  There is nothing like putting in a new system and then 2 months later having to ask for more money to update or change something that was planned incorrectly.  

I think we have decided to have one subnet with the first 254 addresses dedicated to the static devices and the second 254 for the DHCP.  We can merge both locations into one subnet instead of the two separate that we have now.  This should allow us enough space for both with room for growth as well as ease of administration.

So, while not neccessarily going with 'best practice' hopefully we have gotten at least 'good practice' and ease of administration.  A trade-off I am willing to accept :-)

I appreciate your help!
Linds462
0
Question has a verified solution.

Are you are experiencing a similar issue? Get a personalized answer when you ask a related question.

Have a better answer? Share it in a comment.

All Courses

From novice to tech pro — start learning today.