Title: Re: [ActiveDir] In search for duplícate  accounts
I agree with Al ldifde is better for a one-off solution that will search for a few specific entries.  If you really want this type of automation then you should either use some type of DB solution as Al suggests or if your perl is up to snuff use the Perl Net:Ldap module to write in some automation.  A permanent solution to duplicate items would be some sort of Identity Management Solution, like Microsofts MIIS or Novell’s Identity Manager 2.

Brent



From: "Mulnick, Al" <[EMAIL PROTECTED]>
Reply-To: <[EMAIL PROTECTED]>
Date: Thu, 3 Jun 2004 16:36:42 -0400
To: "'[EMAIL PROTECTED]'" <[EMAIL PROTECTED]>
Subject: RE: [ActiveDir] In search for duplícate  accounts

My initial thoughts on this are this:
1) you could do that with a script pretty easily, but that method would be terribly inefficient and cause a lot of traffic (if I understand what you want to do correctly.)
2) a db would be a better suited tool for this task.  Something like access or SQL would be able to find dups based on whatever field you choose.  You would just need to populate the db appropriately.  Access even has the query built in. The advantage here is that you iterate all objects in the forest only once, vs. finding the objects one at a time.  In your psuedo, you have it as "output all samaccountname, sn, and givenname(s) to a file.  Iterate through the file searching on each one for all occurrences and return those to a file".  With that, you'd have a LOT of little files all over the place.  With a DB, you could have the data local and hack and splice until you find the dups pretty easily.

I think changing to csvde vs. ldifde would be easier to import into a db.  It is for me.

My $0.02 (USD) anyway.

Al


From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of "Sanz de León, Juan Carlos"
Sent: Thursday, June 03, 2004 3:22 PM
To: '[EMAIL PROTECTED]'
Subject: RE: [ActiveDir] In search for duplícate accounts

Wow!! Thanks very much for your help Brent.  After your response, some coments that come to mind... maybe you or other LDIFDE experts out there could give me some experiences..

    
How could I have LDFDE OUTPUT the (for example, sAMAccountName,givenName)  and then use it as INPUT into the search you comented below in another LDIFDE command.  In the form of a LOOP.  such that the process would be as follows:

    
a) Output the sAMAccountName, Surname, givenName to console or text file.. -->
   b) and next have LDIFDE input that information and search the GC(3268) for a duplicate in a loop until all users in the forest have been processed.

This would run in a "for loop" until all users in the forest have been completed....( the results would then go to a text file).

I know how to export attribute information from Active Directory using LDIFDE (part a)... what I don't know is how to make it read it in a LOOP until EOF and have it as INPUT into another LDIFDE search.

Or something like that...

Any ideas from anyone out there would be greatly appreciated.
Juan Carlos

    

-----Mensaje original-----
De: Brent Westmoreland  [mailto:[EMAIL PROTECTED]
Enviado el: jueves, 03 de  junio de 2004 17:31
Para:  [EMAIL PROTECTED]
Asunto: Re: [ActiveDir] In search for  duplícate accounts

You will need  to know what values you are trying to find.  For example, people with  duplicate surnames and givenNames or duplicate sAMAccountNames in a forest can  be determined by using ldifde.  The syntax can be a little tricky to the  uninitiated but it is similar to ldapsearch in the unix world.  For  example, if you know the name of the user Foo Bar with givenName Foo and  surname Bar and you wanted to search the forest for all users with that  surname and givenName combination and have the output directed to your console  window, then you would issue the command:

Ldifde -f con -r  "(&(objectCategory=person)(surname=Bar)(givenName=Foo))" -t 3268 -d  "dc=forest,dc=corp" -l "surname, givenName"

Lets step through the  syntax

Ldifde = c:\windows\system32\ldifde.exe if this executable  isn't on your workstation, you should be able to get it off of your win2k  DC.

-f con = the -f switch specifies the output file of the command  and con is console.  So in essence you will be issuing the ldifde command  and directing the output back to your cmd window. You could also specify a  filename if you wanted to dump it into a text file.

-r = The -r switch  indicates the search filter, here you specify the key=value pairs to search  for in the directory. You can & the values; meaning that you can search  for (key=value) and (differentkey=differentvalue).  You can | the  values meaning you can search for (key=value) or  (differentkey=differentValue).  You can also ! The values meaning that  you search for (key=value) but not (differentKey=differentValue).   Search filtering is an art as much as a science and several on this list  can provide a great deal of input on using efficient filters if you are  looking to retrieve specific entries.

-t = 3268 specifies the port  number to contact, because you want to specify all users in the forest it is  best to contact your global catalog.

-d = The searchbase, again if you  want to search for all users in the forest you will need to specify the root  DC entry of the forest.

-l = the -l switch limits the output returned,  without limiting the output of each entry you might return a lot of  information that would be useless.  For example if you are looking for  duplicate surname and givenName combinations, then you probably wouldn't want  to return the exchangeMTA.

You can get more help by typing ldifde  /?

I use it quite often to track down duplicates with great success...  Hope it works for you.

Brent


 

From: "Sanz de León, Juan Carlos"  <[EMAIL PROTECTED]>
Reply-To:  <[EMAIL PROTECTED]>
Date: Thu, 3 Jun 2004  16:50:17 +0200
To: "'[EMAIL PROTECTED]'"  <[EMAIL PROTECTED]>
Subject: [ActiveDir] In search  for duplícate accounts

Greetings gurus,

Does anyone  know of a tool or script that will search the FOREST for duplicate W2k  ACCOUNTS ? We have a forest with about 45 W2K domains... And duplicates are  becoming a problem.

Has anyone ever try to search for duplicates at the  forest Level rather than domain level?

Any tricks to what I want to  accomplish.

Thanks in advance,

JCS

 


Sent using the  Microsoft Entourage 2004 for Mac Test Drive.

Reply via email to