Tuesday, October 26, 2010

SM group project - what we did and how we did it...


Acknowledgments
We would like to thank our SM profesor Dr. Amarjeet Singh for providing us with an opportunity to present a group project. We would also like to thank the creators of wikipedia without which, our project would not have been possible.


We would like to share with you, our experience while preparing this project. Our project was "Data recovery from mistakenly formatted hard-drive" and frankly speaking, when we read this topic in the list of projects, we didn't even know that anything like that was possible. Curiosity is the mother of all creations and the precise reason why all 3 of us, me, Karan Gupta and Mithil Gupta chose this topic.

Since we had no idea what our topic was all about, the first thing that we had to do was find out how is data stored in a hard drive, how is data recovery even possible and how does a utility or a software recover data. Then our task was to gain hands on experience on such utilities and share them with our class.
As far as data storage was concerned, we had been taught in class about file systems and how a file is stored in a hard drive. Still, we went through a number of sites to further strengthen our knowledge of file systems and how they manage data on the hard drive.

Next we had to find out what exactly happens during data recovery or how are various data recovery tools able to recover data from a formatted hard drive. To answer this question, we had to first find out what exactly happens when a drive is formatted. Again, another round of browsing through the net followed. We realised that data is not removed from the clusters of the hard disk but just enabled to be overwritten when new data is to be stored on the hard disk after format. Hence, even if after a disk is formatted, data recovery is possible if the previously stored data has not been over written by newly entered data. We also addressed questions like "is data recovery possible if a partition has been formatted in a different file system than the 1 that existed before? and if yes, how and why?" We learned a lot while making this project and surprisingly, none of us ever felt it like a burden on us. 

The final part of our project involved gaining hands on experience on such utilities and this was the most time consuming and tiring part of the project. There were just way too many utilities. Some links to such utilities are listed as follows:-

PC inspector file recovery. 
Pareto data recovery.  
Disk doctors. 
Ease US. 
Stellar Phoenix data recovery(linux)

We didn’t count but I believe we installed and tried around 10-15 such utilities.  A few didn’t work, others worked but not to our expectations and a few recovered data more or less the way we had expected them to. Above mentioned sites are links to only those that worked really well. There were many utilities which weren’t available for free but we saw their online demos they were much stronger than the ones that we could try for free.

Now the question that arose was that which utility is better than the other, on what basis we should compare and how exactly are they different from each other. The question’s answer lied in the algorithm that each utility used to recover data. There are broadly 2 ways in which data recovery is possible:-
1.      
     One way is to rebuild the data from scratch. This method recovers the files with their original file names. This method of data recovery is usually more useful in case of hard disk crash or any other physical damage to a hard drive but works for a formatted drive as well. It is a very efficient method of data recovery. Hence, most utilities that implemented this type of algorithm weren’t available for free.

2.       An alternate way is that the disc is scanned through, and data is recovered as objects from clusters and linked together. The utilities that we found on the net mainly followed this method. In this method, data is recovered but the original file names are lost. This works well as long as the hard disk has only been formatted and no cluster has been over-written. Moreover, this method is not of any help during hard disk crash or and physical damage to a hard drive.

From a list of more than 10 utilities we chose 3 utilities for our project and did detailed study on them. The main criteria for selecting them was ease of use but since that is relative, we also went through their online reviews and ratings and realized that what we feel about those utilities is also felt by many others. The utilities we chose were PC Inspector file recovery, PareTodata recovery and Stellar Phoenix data recovery, the last one being a tool for data recovery from a linux partition.
Overall, it was a very enlightening experience. We enjoyed it thoroughly and it took our understanding of  a computer’s function of data storage to a whole new level.

Sushant Mehta
2010088

Karan Gupta
2010037

Mithil Gupta
2010049