Take the 2-minute tour ×
Super User is a question and answer site for computer enthusiasts and power users. It's 100% free, no registration required.

Generally questions about virtualization means putting more virtual computers on 1 physical computer. What my question is about is the other way around. Namely: how to make from different physical machines 1 virtual computer?

The point is that I would like it to have the following features:

  • data redundancy: data should be stored on at least x other physical computers

  • access should be able to done on any physical node (so no master node)

  • I would need to be able to place a normal Linux system on top of the virtual layer, so I can get all the benefits of Linux.

  • I would like data to migrate automatically and physical computers to be able to be pulled/added at will.

I've been looking through Google for those things but I didn't find anything that only remotely resembled what I needed. Maybe you can give me a hint where to look at, or on what terms I would need to search?

share|improve this question

migrated from stackoverflow.com Sep 13 '10 at 5:40

This question came from our site for professional and enthusiast programmers.

add comment

3 Answers 3

Potential terms/links that might help narrow your search:

share|improve this answer
add comment

Although not exactly what you describe, have you done any Google searches for clusters or clustering solutions? A cluster doesn't really have a single OS running over the top of all systems, but they are built in a way to transparently share work or data in some manner. It is probably as close as you'll realistically get to what you describe.

Take a look at the NoSQL movement, as there is lots of related activity in regards to migrating data automatically with the ability to remove and add systems at will. You may want to read up on the CAP theorem, here's a reference I found: http://blog.nahurst.com/visual-guide-to-nosql-systems

All of this can get quite complex, and there is no single best solution. As always, you need to find the solution that works best for your use case.

share|improve this answer
add comment

What you are asking for is a global namespace cluster. On top of that you would also like to replicate data across different nodes in the cluster. And you would like to do all of this using Linux. Here's what you are looking for:

Lustre Cluster Filesystem Read more here.
DRBD - Replication across network

And they can both work together, but you'll need to do some investigations regarding performance.

share|improve this answer
add comment

Your Answer

 
discard

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