The Gossip Protocol in Cloud Computing

What is Gossip Protocol?

Gossip protocol is a computer peer-to-peer communication mechanism or process based on how epidemics spread. To ensure that data is distributed to all group members, several distributed systems use peerto-peer gossip. Since some ad-hoc networks lack a central registry, the only method to disseminate shared data is for each member to pass it on to their neighbors.

Because gossip conveys information just like how a virus spreads in a biological community, the phrase "epidemic protocol" is frequently used interchangeably with the gossip protocol.

Periodic, paired inter-process interactions are at the heart of the protocol. The amount of data transferred during these encounters is limited. The information transmitted has an implied redundancy as a result of the duplication. It is intended that this will be used in the future to improve communication.

Types of Gossip Protocols

Dissemination Protocols or Rumor-Mongering Protocols

Dissemination protocols distribute information by chatter; they essentially work by flooding the agents in the network, but in a way that yields bounded worst-case loads.

To carry out multicasts, event dissemination systems use gossip. They report on happenings, yet gossip happens regularly and is not triggered by events. One point of worry is the potential for a long delay between when an event occurs and communicated.

Background data dissemination protocols are constantly squabbling about information related to the nodes that are participating. Propagation latency is usually not an issue, maybe because the data in question changes slowly or because there is no significant penalty for acting on slightly stale data.

Protocols that Compute Aggregates

These generate a network-wide aggregate by sampling the data from network nodes and aggregating the results to produce a system-wide value – the highest value for some measures nodes is taking, the lowest value, and so on.

The aggregate must be computed using fixed-size paired information exchanges; these usually end after several rounds of information exchange logarithmic in system size, at which point an all-to-all information flow pattern has been established.

As a result of aggregation, it is feasible to use gossip to solve different types of problems.

Gossip Protocol in Cloud Computing

The Gossip protocol is used to fix the issues caused by multicasting; it is a method of communication in which a piece of information, or gossip, in this case, is conveyed from one or more nodes to a group of other nodes in a network.

When a group of customers in the network demand the same data simultaneously, this is advantageous. However, other issues arise during multicasting. For example, if there are many nodes at the recipient end, latency (the average time for a receiver to receive a multicast) increases. The gossip protocol sends out the gossip periodically to random nodes in the network to get this multicast message or gossip across the targeted targets in the group. Once a random node receives the gossip, it is considered to be infected by the gossip.

The gossip is received by a random node, which does the same thing as the sender and transmits numerous copies of the gossip to arbitrary destinations. This process continues until the target nodes receive the multicast. After spreading the gossip to random nodes, this method turns infected nodes into uninfected nodes.