Exchange Server, Windows Server, ISA, SEO, Photography

Comments : 8

Finding FSMO Roles in Active Directory using NTDSUtil

Posted by Vishal Vasu | Posted on 04-09-2009

Tags: Active Directory, Windows Server 2003

First things first: FSMO means Flexible Single Master Operation and it is used within Active Directory to control, monitor and manage configuration updates. Without going in to more details about how FSMO works, let’s move on to understanding the Five (5) FSMO roles:

Schema Master: It controls all the updates and modifications that relate to schema of the Active Directory. It is the duty of the Schema Master to ensure that the latest schema has been replicated across all Domain Controllers in the directory. The role is forest-wide and there can be only one Schema Master in a forest.

Domain Naming Master: It controls the addition and deletion of domains in the forest. The role is once again forest-wide and there can be only one Domain Naming Master in a forest.

RID Master: Stands for Relative ID and is responsible for processing RID requests from all Domain Controllers in a particular domain. For example, if we create a user in Active Directory, a unique Security Identifier (SID) is created which identifies it in the domain. Along with this SID a Relative ID (RID) that is unique to the SID is also generated from a pool that is allocated. The main role of RID Master is to ensure that each DC has got enough RID pool and if not then process the request for the pool from the unallocated RID pool. The role is Domain specific and there can be only one Domain Controller handling this role.

Infrastructure Master: When an object in one domain is referenced by another object in another domain, it represents the reference by the GUID, the SID (for references to security principals), and the DN of the object being referenced. The infrastructure FSMO role holder is the DC responsible for updating an object’s SID and distinguished name in a cross-domain object reference. The role is Domain specific and there can be only one Domain Controller handling this role.

PDC Emulator: It seems to be a backward compatible role that has been handed over as a legacy from Windows NT 4.0 and is mainly used for syncing time in the Enterprise. The W32Time service that we see in the Services MMC has a very important role to play: synchronize time across the enterprise to that Kerberos Authentication works smoothly. After all, password changes, authentication, account lockout, etc. all depends on this. The role is Domain specific and there can be only one Domain Controller handling this role.

So why does an administrator need to know which Domain Controller is holding which role? Well, normally there is always a single machine that carries all the five FSMO roles, but there can be scenarios where an administrator would have moved one or more FSMO roles. To know why, read this KB article from Microsoft. With a better understanding of which server holds which role, the administrator can plan better for disaster recovery or scheduled maintenance. There are many ways to find out which DC holds which role, but here is one that uses NTDSUtil – a command line tool.

A word of caution before we proceed: If you do not know what you are doing and have no idea about how Active Directory works or even worse – what Active Directory is – please refrain from doing this practically. Doing this incorrectly or messing with this command can cause your Active Directory to crash and lose functionality.

Step #1: On any Domain Controller, click Start. In the Run command type CMD and hit Enter. You will be taken to the good old command prompt window (DOS were the days). Type ntdsutil and hit Enter.

Step #2: You shall see the screen with ntdsutil: prompt. Since we want to find out the roles, type roles and hit Enter. Notice that the prompt now changes to show fsmo maintenance:. Now is a good time to get more HELP on the list of available commands.

Step #3: On the fsmo maintenance: prompt, type ? and hit Enter. Right-click in the Window, mark and copy them. Paste the clipboard in to Notepad for easy reference.

Step #4: Type connection and press Enter. This will show a prompt with server connections:. Type connect to server <servername> (replace <servername with actual name> and press Enter.

Step #5: Once we are connected to the Domain Controller, type q to return back to the fsmo maintenance prompt. Now type, select operation target and then press Enter. Notice that the prompt changes to select operation target:.

Step #6: At the select operation target prompt, type list roles for connected server and press Enter. This would list all the FSMO roles for that Domain Controller. To get out of the ntdsutil, type q until you are back to the good old DOS prompt.

For those who are looking at ready solution rather than going through this process, check out DUMPFSMOS.CMD in the Resource Kit (Windows 2000/2003). This tool is a one-click utility that does the same trick.

Comments : 9

Open Source versus Open Standards

Posted by Vishal Vasu | Posted on 21-08-2009

Category : General

Tags: Open Source, Open Standards

Everyone’s talking about open source and open standards now-a-days but hardly have I come across any who knows what exactly it contains and what the importance is. I’m putting this post in my blog today as I desire for people to have a better understanding of what is the difference between these two terms. Let’s get a first hand understanding of what these terms mean.

Open Source is a methodology based on which the design, development and the distribution of the developed software is carried out along with its source code. On the other hand, Open Standards means a royalty-free distribution of technology which has been formalized and approved by committees that are open to participation by all interested parties. For example, W3C is a body working on standards for Web Technologies. Click here for more information.

The word “Standards” means a set of guidelines to which a lot of people have agreed upon. Putting this definition in the context of software, “standards” allow a company to pick and choose from competing vendors and interoperate their systems without being pinned down to one of them.

Many think that open source software offers the same advantage, but that is not true. Open source means that you have access to source code for inspection and modifications but it may fail interoperability. This does not mean that open source is not good. There is lots of software available which are better or provide a similar kind of functionality that paid software will do. And the best part of open source is that it is free and comes with the entire source code. But what happens when you need assistance? Is the assistance free or is there a price tag? We all know that there is lot of stake involved in providing support. If you are thinking that support means training people on how to use the product or software, you are mistaken. There is more to support then that. It involves product trainings, more research to fix bugs and vulnerabilities, distributing patches and updates, etc. There may be companies that provide support on open source platforms and have no tie-up or relation with the parent company that designed and distributed the software for free. They may be able to fix small issues since there is access to the source code, but imagine a situation where there is something wrong at the core level and requires dependency on the original developer. Further to this, there is plenty of evidence that suggests that open source does not necessarily equate to cost savings. There are still licensing fees, service fees, implementation fees and support fees associated with it.

I would feel comfortable if I knew that the open source software that I downloaded for free gave me the source code and so I can fix or modify the application as per my needs. But I would be more comfortable thinking that the paid software I use from the vendor is going to be in the industry for long and many brains are working on the software to ensure that it runs smoothly and has improvised functionality with every new release.

All this does not mean that open source is a big NO-NO. Open source does provide the benefit of using someone’s work in your own and then redistributing your work in similar fashion so that someone else can get benefit from your work. This way the community can make progress using a collaborative manner of production. Most of the time this needs to be shared and released under GPL (General Public License) which means that you cannot charge others for a license and you must make your own source code available to the public. This makes it IP (Intellectual Property) unfriendly. Many a times this code will not even make its way to the main stream and will simply end up as an individual company’s product. This leaves a question in mind – will everything in the long run be compatible or provide interoperability? Will my investment in one segment ensure returns in the other or vice-versa? Who’s going to provide updates and security patches to fix vulnerabilities? Who’s going to invest billions to innovate? From my perspective, it’s not a fight between open source and commercial software. It’s all about open standards versus proprietary standards.

I feel we should have more specific and beneficial standards that are not vendor specific or not vendor dictated because ultimately it is the interoperability that counts at the end of the day. If open source software fits your environment and gets the work done in terms of costs, features, support or maintenance – all’s well. But if you are putting security, compliance, performance, upgrades and scalability before everything else then proprietary software designed with open standards in mind is your choice. We can even extend this further and run a combination of both – it’s our choice.

What do you think? Let me know.

Comments : 2

Designing Green Datacenters

Posted by Vishal Vasu | Posted on 24-07-2009

Category : General

Tags: Data Center

Until the knock of environment friendly concepts and public concerns, traditional datacenters had fixed their eye on maximum uptime. Times have changed and datacenters are concentrating on conserving energy. Here are some tips that can contribute towards building eco-friendly datacenters.

Recently, I was involved in redesigning our company’s datacenter to consolidate hardware and systems running at two different physical locations. This was planned not as a “marketing” or “great idea!” move, but in fact it was a pragmatic business decision that helped the company to save money each year. I would like to share here my key learning’s that helped us make decisions to work towards a Greener Datacenter.

Evaluate Current Infrastructure: This is the first step where analysis needs to be done on the energy bills, current consumptions, and how much is being utilized by the servers, air conditioning systems, cooling fans and other devices like routers and switches in the datacenter. This helps in creating a baseline for calculating the Return-On-Investment after the changes are actually put in to implementation. Here is a link to couple of Power Calculators:

India – http://www.indiaenergyportal.org/
UK – http://www.ukpower.co.uk/running-costs-elec.asp
USA – http://www.42u.com/efficiency/energy-efficiency-calculator.htm

Rack Placements: This is an important aspect while designing the datacenter floor plan. Plan the layout of the equipments and structural components in a hot aisle and cold aisle rack layout method as recommended by Green Grid. The basic principle is to maximize the separation between the exhaust air flow and intake air flow. The figure below would give a better idea of the basic design principle.

The benefit of this arrangement is that it affects the required air delivery temperature that must be used to equalize the temperature throughout the room.

Need Analysis: Evaluate what is required out of the datacenter, not just what is needed now but maybe two or three years down the line. This helps in estimating the room for growth and helps avoid costly mistakes which fail to allow enough headroom for IT growth.

Eliminate: Look at every single area where you can eliminate, reduce or consolidate current hardware and equipments. For example, if you have two 24 port switches using only 18 connections consolidate them and shut down one of them. This does not have much impact in small organizations, but when we talk about companies which have lot of equipments in their datacenter, I’m sure figures would be impressive. Moreover, while calculating the ROI also consider the capital equipment and ongoing operational cost savings.

Consolidation of Servers: Identify areas where servers can be consolidated on to one single piece of hardware rather than running multiple physical boxes. By running multiple platforms and applications on a single server with many virtual servers can help reduce the amount of physical boxes. This in turn would help save the power consumption of not only the hardware box but also affect the cooling systems energy consumption.

Usage Patterns: There is a misconception that servers in a datacenter should be available and performing 24×7 regardless of the usage levels. In fact, identify and list out how all different parts of the network are being utilized and what are the peak hours for bandwidth, processing power and storage to handle peak network traffic. Set the power savings features on the hardware or OS level such that it helps in saving power. Allow the servers to sleep when they are not being utilized and explore the Wake-On-LAN technology.

Rated Appliances: Use Energy Star rated appliances, specially the air-conditioning or cooling systems. Devices carrying the Energy Star logo tend to save about 20% to 30% of energy on an average and the specifications differ with each item. Keep an eye on this point as it is estimated that EPA (Environmental Protection Agency) will announce and enforce the ratings for datacenter appliances and servers by 2010. At times, devices with Energy Star rating come costly but it is worth investing that marginal amount during the initial purchase. The overall running costs of power consumption would be much higher if the devices are not power efficient and you will run spending more on the energy costs in the long run.

Monitoring: It’s important to keep a track on the uptime and usage patterns of the datacenter. It basically allows the IT managers to accurately measure the effectiveness of the changes. Moreover, it also assists the top level management to decide on green energy solutions to reduce or eliminate their carbon footprints.

Remember, Green is just not a color – it’s more than that.

Do feel free to add in your comments, ideas and experiences. They would be much appreciated by all those who are planning to make computing energy efficient.