Logstash slowlog field took_in_millis shows microseconds instead of milliseconds

nick4eva · December 28, 2016, 2:20pm

I configured Logstash slowlog like this:

slowlog.threshold.warn: 500ms
slowlog.threshold.info: 400ms
slowlog.threshold.debug: 300ms
slowlog.threshold.trace: 100ms

But I see in slowlog file trace entry with "took_in_millis":103180
I don't believe that my event processing 103 seconds, so I looked at source code and found that took_in_millis calculated incorrectly, it should be took_in_nanos / 1000000 instead of took_in_nanos / 1000

github.com

elastic/logstash/blob/master/logstash-core/lib/logstash/logging/logger.rb#L109


  return  LoggerContext.getContext(false)
end


# Clone of org.apache.logging.log4j.core.config.Configurator.setLevel(), but ensure the proper context is used
def self.set_level(_level, path)
  configuration =  get_logging_context.getConfiguration()
  level = Level.valueOf(_level)
  if path.nil? || path.strip.empty?
    root_logger = configuration.getRootLogger()
    if root_logger.getLevel() != level
      root_logger.setLevel(level)
      get_logging_context.updateLoggers()
    end
  else
    package_logger = configuration.getLoggerConfig(path)
    if package_logger.name != path #no package logger found
      configuration.addLogger(path, LoggerConfig.new(path, level, true))
      get_logging_context.updateLoggers()
    elsif package_logger.getLevel() != level
      package_logger.setLevel(level)
      get_logging_context.updateLoggers()

warkolm · December 29, 2016, 10:28pm

Can you raise that as an issue on Github?

nick4eva · January 3, 2017, 9:44pm

Raised an issue https://github.com/elastic/logstash/issues/6476

system · January 31, 2017, 9:44pm

This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.