Pipeline UI viewer shows zero events everywhere

Elastic Stack Kibana
#1

At first I thought it was due to legacy monitoring still active, I switched to metricbeats in any case, but it still does not show the events in the processing pipeline.

image

The overview screen shows received and emitted events counts alright.

image

ES,LS,KI 7.10

UPDATE

Looks like the issue is the following:

  • We have two nodes running logstash, usually the second one is pretty much inactive and exists for HA purposes
  • Pipeline on both nodes is called "main"
  • When I go to Stack Monitoring -> Logstash -> Nodes -> Node 1 -> Pipelines -> main, and compare that URL to the one of Node 2 it is the same.
    This leads me to the conclusion that only the pipeline of Node 2 is ever shown, for which it is correct that there are 0 e/s as explained above.
    Is this a bug?

(Workaround: I have renamed the pipelines on both nodes to be unique)

#2

Hi @holobolo0815 ,

I'm not quite sure if it's a bug, as the id is the unique identifier for the pipeline. If there are two of the same id, the UI isn't designed to handle that. Perhaps worth filing an enhancement request. It might be helpful to have an out-of-the-box alert for this to let the user know they should change one of the ids

#3

I am not sure what is unique here and what not, in any case: When in both logstash instances the pipeline is called "main", then the link to pipeline stats looks like this both times:

http://kibana.example.com:5601/app/monitoring#/logstash/pipelines/main?_g=(cluster_uuid:n9u_qiS1S9-abCdEfgHijk,refreshInterval:(pause:!f,value:10000),time:(from:now-1h,to:now))

Obviously, when I rename the pipeline, one link becomes

http://kibana.example.com:5601/app/monitoring#/logstash/pipelines/main01?_g=(cluster_uuid:n9u_qiS1S9-abCdEfgHijk,refreshInterval:(pause:!f,value:10000),time:(from:now-1h,to:now))

and the other: http://kibana.example.com:5601/app/monitoring#/logstash/pipelines/main02?_g=(cluster_uuid:n9u_qiS1S9-abCdEfgHijk,refreshInterval:(pause:!f,value:10000),time:(from:now-1h,to:now))

Now I have no idea what cluster_uuid is and how it is related since logstash doesn't have a notion of cluster AFAIK.

#4

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