A simple Lambda to allow clients to retrieve values from Hbase in EMR that have been written by Kafka2Hbase (see https://github.com/dwp/kafka-to-hbase).

To request a record, call the lambda with a document containing a topic and message key from Kafka. The value of that message as persisted in Hbase will be returned.

{
  "topic": "db.claimants.address",
  "key": "1c2c47d6-5b8b-4d82-929e-b5b05b1b2306"
}

If a specific version (other than "latest") is required, specify the version as a numeric timestamp.

{
  "topic": "db.claimants.address",
  "key": "1c2c47d6-5b8b-4d82-929e-b5b05b1b2306",
  "timestamp": 1563809926831
}

The returned value is the raw bytes as received from Kafka. If the integration response is set to convert to text, it will be a base64 encoded string.

Configuration of the Lambda is done entirely via environment variables.

• ZOOKEEPER_PARENT - set hbase.zookeeper.parent to hbase parent URI (default /hbase, aws /hbase-unsecure)
• ZOOKEEPER_QUORUM - set hbase.zookeeper.quorum to a list of hosts (default zookeeper)
• ZOOKEEPER_PORT - set hbase.zookeeper.port to an integer port (default 2181)
• TABLE - the fully qualified Hbase table to read from
• FAMILY - the column family to search for

Rows are retrieved from the configured table using the requested key. The cell value is fetched from the configured column family using the requested topic as the qualifier. If a timestamp is specified, the matching version is returned.

If no valid row is found, null is returned.