Deserialize and preview data from Kafka in DSP

When you use the Kafka source function to ingest data from an Apache Kafka or Confluent Kafka topic, the payloads of the incoming Kafka records are stored in a bytes field named value. During data previews, Splunk Data Stream Processor displays the contents of bytes fields as base64-encoded values. To view the data as human-readable strings during data preview, you must deserialize the data.

Deserializing the value field also makes it usable as input in a wider variety of streaming functions, since most streaming functions do not accept bytes data as input. See the Function Reference manual for information about the data type that each function accepts as input.

Prerequisites

To ingest data from Kafka into a DSP pipeline, you must have a connection to a Kafka broker. See the following pages for more information:

• Create a SASL-authenticated DSP connection to Kafka
• Create an SSL-authenticated DSP connection to Kafka
• Create an unauthenticated DSP connection to Kafka

Steps

1. In DSP, select the Pipelines page.
2. On the Pipelines page, click Create Pipeline.
3. Select Kafka.
4. Configure the Kafka function to use your Kafka connection and get data from your Kafka topic. See Get data from Kafka in the Function Reference manual.
5. On the pipeline canvas, click the Connect a processing or a sink function icon () and then select Eval from the function picker.
6. On the View Configurations tab, enter the following SPL2 expression in the Function field:

   value=to_string(value)
   

7. Click the Start Preview icon () and click the Eval function on the pipeline canvas to confirm that the data in the value field has been deserialized from bytes into strings.