Starting with Logstash 2.3, you can set Logstash to detect and reload configuration changes automatically.
To enable automatic config reloading, start Logstash with the
command-line option specified. For example:
bin/logstash –f apache.config --config.reload.automatic
--config.reload.automatic option is not available when you specify the
-e flag to pass
in configuration settings from the command-line.
By default, Logstash checks for configuration changes every 3 seconds. To change this interval,
--config.reload.interval <seconds> option, where
seconds specifies how often Logstash
checks the config files for changes.
If Logstash is already running without auto-reload enabled, you can force Logstash to reload the config file and restart the pipeline by sending a SIGHUP (signal hangup) to the process running Logstash. For example:
kill -1 14175
Where 14175 is the ID of the process running Logstash.
When Logstash detects a change in a config file, it stops the current pipeline by stopping all inputs, and it attempts to create a new pipeline that uses the updated configuration. After validating the syntax of the new configuration, Logstash verifies that all inputs and outputs can be initialized (for example, that all required ports are open). If the checks are successful, Logstash swaps the existing pipeline with the new pipeline. If the checks fail, the old pipeline continues to function, and the errors are propagated to the console.
During automatic config reloading, the JVM is not restarted. The creating and swapping of pipelines all happens within the same process.
Changes to grok pattern files are also reloaded, but only when a change in the config file triggers a reload (or the pipeline is restarted).