I know most of you guys would love to see messy things getting rectified automatically right?
But hey stop! Do you know how much of effort it has to go into to make a robotic model. Yeah you got it right “a lot”.
So, ask me how did we achieve this or more precisely what we are working on right now. Well read the attached Thesis Paper.
We started working on this prototype since quite some time now but did not publish this paper until now. Got to see similar concepts flying there and thought to publish this paper so that people can understand what we are doing here and what they should expect.
A little snippet of the problem:-
In the era of public, private and hybrid clouds, it is a challenge to minimize the downtime. Our datacenters are no longer a collection of just physical resources connected together but a complex and intricate combination of virtual and physical infrastructure working in perfect harmony with multiple levels of abstraction. As beautiful as this may sound, this whole ecosystem hangs by a very thin thread where a small error threatens to bring down the complete cloud. Finding and fixing problems fast is the key.
Bezoar can automatically understand your cloud architecture, monitor its components, identify root cause, perform impact analysis and heal your cloud. It does not stop there; it is a self-learning system with intelligence to prevent the same problem from recurring in future.
If you would like to know how we have resolved the problem you should read this paper.