Active Directory limits

re: Active Directory limits

bpuhl — Sun, 15 Oct 2006 09:56:13 GMT

For more info on AD limit's, check out Eric's DIT blog at: http://blogs.technet.com/efleis/archive/2006/06/08/434255.aspx

Active Directory limits

Santhosh Sivarajan — Wed, 18 Oct 2006 20:35:31 GMT

nice blog.. http://blogs.technet.com/mcs-ireland-infrastructure/archive/2006/10/14/active-directory-limits.aspx...

Limitele Active Directory

Andrei's Weblog — Mon, 23 Oct 2006 14:30:11 GMT

Daca sunteti curiosi care sunt limitele teoretice ale resurselor intr-un mediu Active Directory puteti...

re: Active Directory limits

grillenmeier — Tue, 31 Oct 2006 14:19:28 GMT

this is a good list, but you may want to differentiate between "hard" and "soft" limits. For example, even in Windows 2000 the maximum number of group members is not set to a hard limit of 5000 (you can have plenty more - I've seen group with 24.000 members in Win2k), however once you go beyond the recommended limit you risk replication issues in AD. That's why this "soft" limit of 5000 members to a group basically still exists in Win2003 - it's just not regarding the total number of members in a group, but you should also not add or remove more than 5000 members at once in Win2003 to avoid replication issues.

Along this line, it would be great if you could add the different sources for your limits (which is where we would likely find more inforamtion about the reasoning for these restrictions).

re: Active Directory limits

mcsieinf — Wed, 01 Nov 2006 23:49:26 GMT

The difference between hard and soft is irrelevant if the limits are those that are "supported". There is no difference between something the product will not do, and which it will potentially do if either scenario is not supported by Microsoft.

We will work on identifying the sources for each in another post - that's a great idea.

re: Active Directory limits

mcsieinf — Wed, 01 Nov 2006 23:52:43 GMT

In terms of the "soft" limit of 5k members per group - this is in terms of the impact due to the replication mechanism in w2k versus w2k3. In w2k the limit was indeed soft, but exceeding it could lead to undesirable replication overhead. With the change in replication in w2k3, this soft limit was no longer an issue because only the deltas would get replicated as opposed to the entire group membership.

Of course, if you add a large number of members to a group at once you will get replication overhead regardless as you say - however the number of real world scenarios where this would happen is very small - apart from perhaps a migration context which in itself should be treated as a special case from an operational perspective.