Theoretical foundations of robust and distributed services

I have the notion of a social network which is robust against malicious attacks from the outside. My vision is a system that is structurally built up as a distributed network of equal servers that operate on the same data and offer the identical services. Users should all interact on the same network, independent of the respective server they communicate with. The idea is to prevent (at least theoretically) every opportunity to bring the whole system down and to achieve a total failure of the whole service. Obviously the design of such a system is not trivial because of proper synchronization mechanisms. On the other hand there is the challenge to convey the participants to the appropriate servers. Hope you got my idea roughly.

The thing is that I only have a vague notion how such a system can function, and furthermore have seemingly successful passed over the relevant lectures about distributed systems in my cs studies. I'm therefore a little bit in a lack of an overview of relevant literature, scientific papers describing theoretical models and also real world examples.

Can anyone help me out with references, links and helpful explanations?

You might be interested in these papers:

-
The LOCKSS system

-
Publius: A robust, tamper-evident, censorship-resistant web publishing system

-
The Eternity Service

-
The Free Haven Project: Distributed Anonymous Storage Service

-
Freenet: A Distributed Anonymous Information Storage and Retrieval System

-
Tarzan: A Peer-to-Peer Anonymizing Network

-
Anonymizing Censorship Resistant Systems

-
Tangler: A Censorship Resistant Publishing System Based on Document Entanglement

-
The Economics of Censorship Resistance

StackExchange Computer Science Q#13261, answer score: 5