Filters as processors
You can use filters as processors in Fluent Bit. Placement restrictions depend on the filter: for example, in v5.2 or later, Rewrite tag supports only input processors and rejects output processor placement. Earlier versions accepted it as an output processor, but that placement wasn't reliable and shouldn't be used.
Examples
The following examples show how to configure filters as processors.
Grep
In this example, the Grep filter is used as an input processor to filter log events based on a regular expression pattern:
parsers:
- name: json
format: json
pipeline:
inputs:
- name: tail
path: /var/log/example.log
parser: json
processors:
logs:
- name: grep
regex: log aa
outputs:
- name: stdout
match: '*'Lua
In this example configuration, an input plugin uses the Lua filter as a processor to add a new key hostname with the value monox. Then, an output plugin adds a new key named output with the value new data.
Rewrite tag
Use Rewrite tag as an input processor to route records under a new tag and continue at the next processor stage of the same input. Its experimental batch_size option can group records by destination tag to reduce downstream processing overhead. Batching is disabled by default and can change interleaving between different destination tags.
Last updated
Was this helpful?