Tuesday, March 08, 2005

Exchange Tidbits

Evan Dodds posted recently a few answers to common questions.  Here are a few of them:

Couple of Exchange tidbits I’ve been saving up for a while. None of them are on my usual topics (Exchange clustering, site consolidations, etc) and they are not large enough individually to warrant their own post, but hopefully they’ll help you out.

1) You can force the Exchange 2003 SP1 RUS to stamp secondary email proxy addresses onto an existing user account without having to do “apply now” on the recipient policy. This is principally for cross-forest migrations, but may also work for users who haven’t been moved cross forest. KB.820381 covers the cross-forest scenario, but you can also set the GUID manually on the user’s msExchPoliciesIncluded attribute: {23668AD4-4FA1-4EE8-B2BB-F94640E8FBA0}.

2) NTDSNoMatch doesn’t have to be set on Custom-Attribute-10. Everyone refers to extensionAttribute10 when discussing NTDSNoMatch (see KB.274173, for instance). Little-known fact is that it doesn’t have to be attribute #10. If you’re using attribute #10 for something else, just put the “NTDSNoMatch” value into any of the 15 extensionAttributes and the ADC will pick this up transparently. Thanks to Alex Seigler for this find.

3) LegacyExchangeDN may add a random 8 digit number to ensure uniqueness. You may notice that sometimes the LegacyExchangeDN value associated with a user is of the format /o=org/ou=site/cn=recipients/cn=username######## rather than the expected /o=org/ou=site/cn=recipients/cn=username. This happens because the LegacyExchangeDN serves as the unique X500 identifier in the Exchange 5.5 directory, and as such, has to be unique. If there is any collision of LegacyExchangeDN values when the user is being mailbox enabled, eight random numeric digits are appended to the proposed LegacyExchangeDN and it’s checked again for uniqueness. Note that it really doesn’t matter what your LegacyExchangeDN is, so long as it’s properly associated with the user and is unique (ie - don’t bother trying to go back and “change” these unique-ified values)

No comments: