Skip to main content

stdin

Consumes data piped to stdin, chopping it into individual messages according to the specified scanner.

# Config fields, showing default values
input:
  label: ""
  stdin:
    scanner:
      lines: {}
    auto_replay_nacks: true

Fields​

scanner​

The scanner by which the stream of bytes consumed will be broken out into individual messages. Scanners are useful for processing large sources of data without holding the entirety of it within memory. For example, the csv scanner allows you to process individual CSV rows without loading the entire CSV file in memory at once.

Type: scanner
Default: {"lines":{}}
Requires version 1.0.0 or newer

auto_replay_nacks​

Whether messages that are rejected (nacked) at the output level should be automatically replayed indefinitely, eventually resulting in back pressure if the cause of the rejections is persistent. If set to false these messages will instead be deleted. Disabling auto replays can greatly improve memory efficiency of high throughput streams as the original shape of the data can be discarded immediately upon consumption and mutation.

Type: bool
Default: true