Thanks for the thoughtful comment.

I am aware of page sizing when retrieving result from AD. This isn’t the cause of problems with the second example as I am using a very small test domain in VMware to sandbox my testing from live. I’m glad you raised the point though. I had been deliberatly keeping my examples simple.

I have read a couple of developers rating the book you mention highly. I should really add it to my Amazon Wishlist.

While we’re on the topic of books, if anyone is interested in good AD books then the two I highly recemend are:

Active Directory Forestry
by John Craddock and Sally Storey
# ISBN-10: 0954421809
# ISBN-13: 978-0954421809

Active Directory Cookbook for Windows Server 2003 and Windows 2000
By Robie Allen
# ISBN-10: 0596004648
# ISBN-13: 978-0596004644

They are no related to PowerShell, howerver the AD content of both is excellent!

Cheers

Adam

Thanks for the posts! I have a couple of comments on the Directory Searcher samples you show in this post:

Your first sample is the correct way to search. It’s ALWAYS more efficient to filter in the provider (whether you’re using directorySearcher() or get-childitem – whatever). When you do a findAll in the second sample you may be trying to retrieve 1000′s of objects in any reasonably sized directory.

Secondly, this is probably the reason your second sample gives inconsistent results. The search will only return a fixed amount of information in one chunk – even if more items match your query. If the entry you need happens to be in the chunk returned then your code will work. If it’s not then the code will fail. The way around this is to do a paged query (set the pagezise attribute to “1000″ – the actual value is fairly irrelevent, but 1000 is the value used by convention). This will ensure you get ALL the results back (could be SLOW!). Alternatively – and preferably – you can get around this by filtering in the provider as in the first sample. [I've just read your Search Tips and you allude to this there:-)]

Thirdly (a small point) – findone() under the covers actually does a findall() and returns the first result. I find this misleading so I always explicitly use findall() because it’s clear what’s going on. You might think findone() is quicker, but actually it’s not!

Finally, you’re right – it’s more functional to return a directoryEntry rather than a searchresult – but retrieving the directoryEntry always results in ANOTHER LDAP query – so if efficiency is important you should only call GetDirectoryEntry() when you really need too.

HTH!

Chris

PS. I only know this stuff because I read “The .NET Developer’s Guide to Directory Services Programming” by Kaplan & Dunn. An excellent book which I recommend to people whenver I get the chance!