The name of the Kinesis stream to consume
The configuration
to use for processing records - this is passed to, and used by, the underlying Java Amazon library. Defaults to Executors.newCachedThreadPool()
A function that is called for each "batch" of records received from the kinesis stream. This function may be called from different threads (for different shards), and possibly even concurrently. Note this function should complete in less time than the configured KinesisConsumerConfig.leaseFailoverTime as otherwise the lease could expire while processing is proceeding.
The configuration
to use for processing records - this is passed to, and used by, the underlying Java Amazon library.
to use for processing records - this is passed to, and used by, the underlying Java Amazon library. Defaults to Executors.newCachedThreadPool()
Register a function to be called when a shard is being terminated.
Register a function to be called when a shard is being terminated.
This can be useful to allow applications decide their own checkpoint workflow, rather than checkpointing every "batch".
If no function is registered here, then an automatic checkpoint is issued on shard termination.
Only a single registration is allowed.
A function that will be called on shard termination - this is passed a Checkpoint and the ID of the shard being terminated.
Initiate the shutdown of processing the stream.
Start reading data from the stream.
The name of the Kinesis stream to consume
A Consumer of data from a Kinesis Stream.
If multiple shards exist, the processBatch function here will be executed from different threads, possibly concurrently.
The name of the Kinesis stream to consume
The configuration
to use for processing records - this is passed to, and used by, the underlying Java Amazon library. Defaults to Executors.newCachedThreadPool()
A function that is called for each "batch" of records received from the kinesis stream. This function may be called from different threads (for different shards), and possibly even concurrently. Note this function should complete in less time than the configured KinesisConsumerConfig.leaseFailoverTime as otherwise the lease could expire while processing is proceeding.