Metricbeat Logstash module payload size


I opened a ticket about this issue in the github, but it got closed and the issue was redirected here:

It is my understanding that the Logstash UDP output plugin doesn't do any batching, it just takes a single event and wraps it into a UDP packet and off it goes, correct?

I think the person who closed the ticket made a wrong assessment about the issue, I can confirm that with either node or node_stats metricset enabled/disabled, the module sends the pipeline representation information which is the problem here, because it is so large (in my case, even though I have only about 10 filters and 1-2 inputs and outputs) that the limit of UDP packet is exceeded.

So based on this information, I would like to open a discussion about enabling Logstash node monitoring without pipeline monitoring. I though it would be already possible by selecting either node or node_stats metricset but that is not the case.

ping @andresrc

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