The publish/subscribe paradigm is
gaining attention for the development of several applications in wide area
networks (WANs) due to its intrinsic time, space, and synchronization
decoupling properties that meet the scalability and asynchrony requirements of
those applications. However, while the communication in a WAN may be affected by
the unpredictable behavior of the network, with messages that can be dropped or
delayed, existing publish/subscribe solutions pay just a little attention to
addressing these issues. On the contrary, applications such as business
intelligence, critical infrastructures, and financial services require delivery
guarantees with strict temporal deadlines. In this paper, we propose a
framework that enforces both reliability and timeliness for publish/subscribe
services over WAN. Specifically, we combine two different approaches:
gossiping, to retrieve missing packets in case of incomplete information, and
network coding, to reduce the number of retransmissions and, consequently, the
latency. We provide an analytical model that describes the information recovery
capabilities of our algorithm and a simulation-based study, taking into account
a real workload from the Air Traffic Control domain, which evidences how the
proposed solution is able to ensure reliable event notification over a WAN
within a reasonable bounded time window.
No comments:
Post a Comment