A Combiner transformation combines multiple events from a streaming source into a single event based on the specified conditions.
A Combiner transformation processes binary data and JSON data. For JSON message formats, the Combiner transformation combines the incoming data into an array of data and returns JSON array objects as output. For binary message formats, it combines the incoming data based on the specified conditions.
In a
streaming ingestion
task, you can add only one Combiner transformation. If the task includes both a Combiner transformation and a Format Converter transformation, the Format Converter transformation must be the last transformation in the task flow, preceded by the Combiner transformation. If the task doesn't include a Format Converter transformation, the Combiner transformation must be the last transformation in the task flow.
You can use one of the following conditions for a Combiner transformation:
Minimum number of events
Maximum aggregate size
Time limit
For example, consider the following events:
Record created
Record published
If you use comma (,) as a delimiter, the Combiner transformation returns the following combined event:
Record created,Record Published
When you process binary data with a Combiner transformation, you cannot use a regular expression as a delimiter.