Reliable communication in the presence of failures
ACM Transactions on Computer Systems (TOCS)
The causal ordering abstraction and a simple way to implement it
Information Processing Letters
Lightweight causal and atomic group multicast
ACM Transactions on Computer Systems (TOCS)
An optimal algorithm for distributed snapshots with causal message ordering
Information Processing Letters
Fault-tolerant broadcasts and related problems
Distributed systems (2nd Ed.)
Time, clocks, and the ordering of events in a distributed system
Communications of the ACM
Genuine atomic multicast in asynchronous distributed systems
Theoretical Computer Science
Distributed Algorithms
A Non-Blocking Lightweight Implementation of Causal Order Message Delivery
Selected Papers from the International Workshop on Theory and Practice in Distributed Systems
The Hierarchical Daisy Architecture for Causal Delivery
ICDCS '97 Proceedings of the 17th International Conference on Distributed Computing Systems (ICDCS '97)
Causal separators for large-scale multicast communication
ICDCS '95 Proceedings of the 15th International Conference on Distributed Computing Systems
Total order broadcast and multicast algorithms: Taxonomy and survey
ACM Computing Surveys (CSUR)
Live linked data: synchronising semantic stores with commutative replicated data types
International Journal of Metadata, Semantics and Ontologies
Hi-index | 0.00 |
We study the fifo and causal multicast problem, two group-communication abstractions that deliver messages in an order consistent with their context. With fifo multicast, the context of a message m at a process p is all messages that were previously multicast by m's sender and addressed to p. Causal multicast extends the notion of context to all messages that are causally linked to m by a chain of multicast and delivery events. We propose multicast algorithms for systems composed of a set of disjoint groups of processes: server racks or data centers. These algorithms offer several desirable properties: (i) the protocols are latency-optimal, (ii) to deliver a message m only m's sender and addressees communicate, (iii) messages can be addressed to any subset of groups, and (iv) these algorithms are highly resilient: an arbitrary number of process failures is tolerated and we only require the network to be quasi-reliable, i.e., a message m is guaranteed to be received only if the sender and receiver of m are always up. To the best of our knowledge, these are the first multicast protocols to offer all of these properties at the same time.