Canal Chapter 8

canal If not implemented HA, It can also run , There is no need for zookeeper. If canal Need to achieve HA, Need to rely on zookeeper, Core configuration such as ：canal.zkServers = master:2181,slave1:2181,slave2:2181, This configuration is explained in the following chapters .

Do real-time calculation , Usually by Spark perhaps Flink To consume in real time kafka Data in , There is no client for real-time computing , Or write Flink Code , Or write Flink SQL.
canal Cluster construction depends on zookeeper colony , If the data needs real-time calculation ,canal The collected data needs to be written into kafka, Then we can do real-time calculation .
there canal The client is just testing ,canal Whether the tool can collect data from the database , In practice ,canal The client may be spark Application or flink application .