So, it’s pretty well understood what Active Directory (AD) is. Since we shipped Windows 2000 several years ago, we’ve been putting information out there on what AD is, how you use it, how it benefits you, and why you should use it for as much integration as you can. I think AD sells itself to a large degree.
More recently, we released a product called Active Directory Application Mode (ADAM). While we have several good “what is ADAM” docs out there, I wanted to take a swing at it and ensure we’re all level-set on what this new directory technology really is. This one is sure to come up again.
I am a huge ADAM fan. I enjoy working with it, talking about it, thinking about it, and just generally evangelizing it.
I felt I had to say this now as the future of this blog will be tainted with my bias. You should know it’s there. :)
So, you might be wondering why I’m so pro-ADAM. What do I think it can do that’s so cool? My reason is actually quite silly, and has little to do with using ADAM in the real world.
You see, I enjoy working with the innards of AD. That is, the core directory itself. Sure I enjoy working with all of the stuff on top of it….authentication/authorization, policy, DNS (or is DNS below it? A debate for another day….), and countless other technologies….but in reality, it is the core directory I enjoy working with most.
So where does ADAM fit in? Well, think of ADAM as the core directory of AD, stripped out, and called its own service. It has many of the LDAP and RPC interfaces we use in AD, but without the deep OS integration we’ve come to expect. One can run ADAM on any member server, domain controller or workstation in your environment (in any domain or workgroup). While we encourage it to be used on Server 2003 when in production, ADAM will also run on Windows XP (with some minor functionality not available).
Of course, my personal interest is probably not enough to inspire many organizations to deploy this product (although perhaps it should be! ;)). The following scenarios tend to be where we see ADAM used most:
These are just a few scenarios of course; we have more outlined in the documents available from the ADAM home page. And others would come up with scenarios of their own too not listed in our documentation.
While we’re on documentation, I’d like to suggest going to the homepage I linked to above, and checking out the Active Directory Application Mode Technical Reference. In terms of depth, that has the most information out there.
I’d be remiss if I didn’t mention the FAQ. In addition to those questions a few more that I have been asked recently:
Q: How does <insert AD component here> work in ADAM? I don’t see any documentation on it.
A: When it comes to inside the directory type concepts, ADAM maps to AD in Server 2003 when you have the forest functional level set to the max (2003 forest functional, aka ==2). So all of the fun things like enhanced schema defunct, linked value replication, increased multivalue limit, etc.
Q: I’ve read quite a bit about the user proxy functionality that is in ADAM. How do I know if that’s for me?A: I’d like you to indulge me, and let me skip this for now. I have a future post just dedicated to this one functionality. The FAQ on the ADAM homepage mentioned this a bit, but we’re going to dedicate some more time to it.
Q: What does the ADAM schema look like compared to the AD schema?
A: During ADAM setup we give you the option of importing several elements that help close the gap between the two schema’s (specifically for InetOrgPerson and User). That said, they are still not identical. There are some items in the AD schema that you will not find in ADAM, but you are more than welcome to migrate them yourself. And I’m happy to help if you’d like it. :)
Q: How well will <insert your application name here> integrate with ADAM?
A: That’s a tough question to answer in the aggregate. Most applications which are pure LDAP-consumers will integrate nicely with minimal work. Some other applications (for example, those that use integrated Windows authentication) may require more work. Still others might not port well at all due to the lack of Windows security infrastructure integration or other AD-specific functionality. As we go through more ADAM-related posts, I think this question will answer itself to a large degree.
Q: I didn’t see mention of pricing. What does ADAM cost?
A: ADAM is licensed just like AD. That is, so long as you have your Windows CALs set up (or however you license Windows server and client today), ADAM is licensed too.
I have quite a few upcoming posts on ADAM in my head, including: high availability deployments, user authentication, schema migration, and performance tuning. These topics make the best ADAM posts (I think) because they are where we differ the most from AD (schema migration perhaps the delta is small, but we do get the question often). Most other topics we’ll cover by virtue of talking about them in AD, and ADAM differences will be noted as such. Please do holler if there are other ADAM-specific things that you think need attention which I didn’t mention.
I don't think AD have any future applications,WinFS is much more flexible and simple to use architecture or simple put relational/XML/object mix.
Well, I can do it all according to some management.
LOL, they were pitching the pros and cons of ADAM to our development director and head team leaders. When it was all said and done, they all asked the same question, how is Adam (referring to me) going to do all this?
Andrey, surely you don't think that the introduction of the WinFS data store will somehow eliminate the need for directories do you?
WinFS, like many other data stores in the past and someday in the future, is a rich store built on top of a relational database. It solves a multitude of problems and brings a lot to the table. But it does not aim to eliminate the need for directory information.
AD has always provided a mechanism by which non-dynamic data could easily and securely be stored in such a way that it is highly available to a multitude of clients, Windows and non-Windows. WinFS can certainly help with this, but it is not a panacea. Thinking that it is seems short sighted to me. It has a different set of design goals from AD, ADAM or other directories out there.
WinFS doesn't aim to solve many things, like a fully integrated user authentication & authorization story. Now, a different way to say this would be "perhaps AD will integrate with WinFS." If you made that statement, I could see your perspective, and maybe we will. But saying that WinFS somehow replaces directories is simply not correct. That's just not the goal.
Well, from my point of view, ADAM is great for writing AD applications. And these applications will stay for some time even after WinFS. I'm happy to have new AD up and running in like 5 minutes after the havoc I arranged with the old one. Or to have the chance to test as crazy script as possible without the fear of loosing my precious big AD. But the problem is that there is no enough noise around it. Not enough whitepapers, cases, presentations or "Do more with less" slogans. And without them it is really hard to convince my managers that this is one great product and we have to invest time in it. Are there any plans to change this situation?
Hi Lazar. I hear what you're saying. ADAM has been put out there, but not with the hype or fanfare that some might hope for, or even expect.
Please do holler as to what would help you. Case studies, I hear you. In terms of white papers....what topics would you like to see covered that you don't today?
Dynamic (meta)data is "MUST have" for future systems but AD is too old techology to catch up with modern requirements of distributed AND dynamic world IMHO.Relational databases can be dynamic from definition to data layer but AD not,so why improve something that dosen't play nicely from start with current systems/demands?You position AD/AM as lightweight web authorization data store or integration="front end layer" for legacy AD based systems(mostly) but I think twice before implement rigid AD infrastructure for my web(smart?) apps(I tried it from AzMan point of view...hard experience=nothing work from start,something like linux;-)
Why so many extranet/internet(truly distributed infrastructure) based on sql/forms/ssl authentification schemes today if AD such a beautiful thing?Ask youself,what's wrong with it?True innovations usually dosen't pass by unnoticed,I know about AD/AM from begining and participated in beta-nothing excited so far,sorry.
Hello Eric, I'm just a simple dev, so no hollering from me. In "Active Directory Application Mode (ADAM) and the ADAM Technical Reference" chat, some possible walkthroughs were mentioned (using ADAM as user store in extranet scenarios for example). What happened with these? Then chat if from May 2004. Another idea which I saw in MS CMS newsgroups was use of ADAM for user authentication in MS CMS 2002 sites. Furthermore some topics in ADAM-SDK.chm file are marked as preliminary (Active Directory Application Mode Reference is good example) and the product was released like a year ago. It may sound bad, but ADAM feels like forgotten product. Thank you for your time.
By a curious coincidence, I've been thinking recently about using ADAM in exactly one of those extranet scenarios - for holding user registration and profile information across multiple web-sites and web services.
Is there any sample code available along those lines? I was considering starting with an ASP.NET 2.0 membership/profile provider to sit over ADAM, but I'd be interested to see any prior are out there.
it discusses how to use ASP.NET 2.0 membership over a custom ADAM schema it includes sample code.