We run our elastic stack utilising ECK however have run into an issue when it comes to Logstash logs.
Our Logstash's are deployed using the official helm chart - Logstash Helm
We expected, as outlined with the Elasticsearch and Kibana nodes that we could install Metricbeat and Filebeat to monitor the container and ship the logs into elastic to show up in stack monitoring. Metricbeat works fine however there seems no support for Filebeat (Monitoring Logstash). There seems to be no Logstash log file at all within the container when exec'd into it.
We assume that all logs are sent directly to stdout or stderr and therefore never write to a log file. Is there anyway we can retrieve these messages and ingest them into the cluster?
On a similar note, we have the same issue with nginx ingress controller which symlinks the error and access log files to stdout and stderr so filebeat sends nothing.
Has anyone got a workaround for this or can add anything else to this issue?
Many Thanks