Kafka - Topics đŋ
Description đ
Topics are a particular stream of data, like a table in a database (without the constraints). You can have as many topics as you want, and is identified by its name. Topics also support any message format. The sequence of messages in a topic is called a data stream. You can not query a topic directly, but you can use a consumer to read the data stream from the topic. Similarly Kafka uses a producer to write the data stream to the topic.
Kafka Topic Durability đĒ
As a rule, for a replication factor of N, you can permanently lose up to N-1 brokers and still recover your data. For example, if you have a replication factor of 3, you can lose 2 brokers and still recover your data. This is because the data is replicated on the remaining brokers. If you lose all the brokers, you will lose all the data. This is why it is important to have a replication factor of > 1.