IT management is the most talked about topic in IT community. Every other company is looking out for some solution that would help them manage their IT infrastructure.
I found that there are some of the major start-up companies that are fighting in IT management space. They are following:

Cendura
Troux
Rellicore
Collation

It would be interesting to watch who becomes the winner in this landscape.

When the question is about managing what? I would say, manage complex distributed application with instances running on 1000 machines! Is that easy? Absolutely Not! But this problem would eventually get solved soon! There is so much to manage in a software application. They are following:

1. Manage all the important files (exe, dlls etc etc)


2. Manage documentation


3. Manage configuration files particularly


4. Manage databases

So what would we do by managing them? How would we manage them?...

Today, when there is so much focus on managing complex distributed applications, the question that arises is what to control? Say, you have oracle database instance running on say 100 machines. And some mirror website went down because of oracle instance going down? So, this demands management of oracle database on all the machines. But, then if you manage oracle database instance, what about managing apache server instances. What if it was discovered that website went down because the server configuration setting was changed? So, we come down to managing/controlling apache server and oracle database instances on all the 100 machines. But how do we do? That is the biggest question mark in current computer world? Is there some solution to it? I would say NO! But are people working to find out the solution? I would say YES. Soon, there would be a market leader in this area.

After troubleshooting for some hours, the technical person found that somebody has played around with configuration settings of one of the file sometime yesterday. That led to the crash of website. But who changed? At what time it was changed? Why changed? And who would be accounted for loss of users + money? The "somebody" or the technical management. Why is not there a solid control on the software infrastructure?

Did you ever came across some website that gave the sorry message for being down? Did you know that if yahoo goes down for one day, how much are they going to loose in terms of money and credibility from user? Thats just a tremendous value. What is amazon goes down for 4 hours? Do you know how much sale are they going to loose?

So what should they do to keep the downtime really low? What should they do to troubleshoot the problem very fast? These are some burning topics of current software world arena! Watch out for further discussion!

So, what could go wrong in an application whose instances are running on large number of machines, say, 1000 machines? Say, a machine in a data center is down. Yesterday it was running great! Has someone changed some configuration parameter remotely? Has some files be removed? Who would know this? Is there a cool way to know this?

Configuration management, change management of distributed applications running on, say 1000 servers is the buzz word in current software products' arena. Who are the players who provide the software to manage distributed applications? How best could the distributed applications be managed? How to reduce the expenditure related to configuration management? These are hot questions of software market in recent times. All the big companies are looking for one or the other products. And, there would be winner in next 12 months. But which company would be winner is there to watch?

Complex distributed applications! Hard to manage them! Downtime leading to huge losses to the companies! I have been working in this area since last few years. So thought of writing my opinion!