Take the 2-minute tour ×
Server Fault is a question and answer site for system and network administrators. It's 100% free, no registration required.

Please suggest me a few research topics ....

My background is from setting up Windows 2003 servers and various firewalls ranging from opensource to commercial ones (both hardware and software).

I also do c# programming...but I don't want to get stuck on a research paper which involves intensive programming.

As of now,I have completed my masters in I.T Thanks

share|improve this question
2  
This should be community wiki, surely? –  ThatGraemeGuy Jul 22 '09 at 11:55
    
What is your PhD in? IT? Computer Science? Business? –  duffbeer703 Jul 22 '09 at 12:38

4 Answers 4

up vote 6 down vote accepted

Coming from a system admin background you could look into self healing networks and algorithms for effective concurrecy, redundancy and scalability.

Maybe look into some protocol/system independent way to setup a network that can self-reconfigure on failure.

For example, if a mail server dies, then due to it's configuration being redundantly backup to all nodes in real-time (regardless of it being a mail-server, this could be filesystem level) then the network can negotiate the best system to redeploy the mail server on.

Algorithms and setups for this will cover alot of difficult subjects such as graph traversal (weighted by performance latency for redeployment), concurrent state, clustering (and issues like split-brain) and tons more.

An application/os/system agnostic approach might be interesting.

You might find alot of the algorithms and setups used in large SAN deployments may help. Translating block-level failure/redundancy and SAN failover into generalized machine state without virtual-machines would be cool. A virtual-machine's state being backed-up is one thing, but often environments run a unified OS and therefore, you don't need all that state.

share|improve this answer
    
A very good suggestion. –  Chris Jul 22 '09 at 10:16
    
If you go down that road, send me a copy of your thesis! :) –  Aiden Bell Jul 22 '09 at 11:30

You could combine your programming and firewall experience into something security related. Algorithms for adaptive threat monitoring at different levels of the OSI model, something like that.

share|improve this answer
    
+1 sounds awesome ... –  Aiden Bell Jul 22 '09 at 18:34

I ran into the problem if remotely upgrading the firmware of 1000 devices. The devices were deployed randomly and with no physical access but were connected to the internet and already had some kind of remote control software available. This produced several questions which could be investigated.

  • How to deliver the firmware without killing the server hosting the file
  • How to track successful upgrades
  • How to create smaller test groups to test the upgrade before deploying it everywhere
  • How to choose members for such groups
  • Can a peer-to-peer deployment mechanism be used instead of central server
  • How to track offline devices upgrade those later
share|improve this answer
4  
If these are PhD topics, the standards for awarding PhDs have really gone downhill. –  womble Jul 22 '09 at 10:25
    
Seriously. Isn't a PhD supposed to represent original research? The valet at my uncle's restaurant solves problems like this. –  duffbeer703 Jul 22 '09 at 12:44

The more important step is to find a Ph.D. program that has ongoing research in IT-related topics (I'm not aware of any, but I haven't looked). I am currently stalled out in my Ph.D. studies, and I have often considered trying to find something in the IT sphere, since that is really where my expertise is. As Aiden suggests, autonomic systems is a field where there is some industry interest and you are likely to find some active researchers.

share|improve this answer

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.