Consumers "pull" messages from Kafka topics. By pulling messages, multiple consumers can consume messages at the maximum rate without being overwhelmed or effecting other consumers reading the same data.
Consumers pull messages in a sequential order. The current position or "offset" records where a consumer is in reading linear messages. If a consumer goes down, it can immediately pick up where it left off based on the saved offset position.
Consumers exist in groups. A single group consists of individual members. The group as a whole reads from a given topic. The number of partitions for a given topic are equally "load balanced" across members within a consumer group.
Consumers "pull" messages from Kafka topics. By pulling messages, multiple consumers can consume messages at the maximum rate without being overwhelmed or effecting other consumers reading the same data.
Consumers pull messages in a sequential order. The current position or "offset" records where a consumer is in reading linear messages. If a consumer goes down, it can immediately pick up where it left off based on the saved offset position.
Consumers exist in groups. A single group consists of individual members. The group as a whole reads from a given topic. The number of partitions for a given topic are equally "load balanced" across members within a consumer group.