sscaling / docker-jmx-prometheus-exporter

Dockerized jmx-exporter for prometheus

Docker JMX exporter for Prometheus

Essentially another dockerised JMX Exporter image, this uses alpine-java and dumb-init to provide a relatively small image (approx 130Mb) and includes a released version of jmx_exporter from the maven central repository

Available on Docker hub

Building docker image

docker build -t sscaling/jmx-prometheus-exporter .

Running

docker run --rm -p "5556:5556" sscaling/jmx-prometheus-exporter

Then you can visit the metrics endpoint: http://127.0.0.1:5556/metrics (assuming docker is running on localhost)

Configuration

By default, the jmx-exporter is configured to monitor it's own metrics (as per the main repo example). However, to provide your own configuration, mount the YAML file as a volume

docker run --rm -p "5556:5556" -v "$PWD/config.yml:/opt/jmx_exporter/config.yml" sscaling/jmx-prometheus-exporter

The configuration options are documented: https://github.com/prometheus/jmx_exporter

Environment variables

Additionally, the following environment variables can be defined

• SERVICE_PORT - what port to run the service (if you don't like 5556)
• JVM_OPTS - any additional options, Xmx etc.
• CONFIG_YML - override the location of config.yaml (default: /opt/jmx_exporter/config.yml which monitors jmx exporter's jvm)

Using with Prometheus

Minimal example config:

global:
 scrape_interval: 10s
 evaluation_interval: 10s
scrape_configs:
 - job_name: 'jmx'
   static_configs:
    - targets:
      - 127.0.0.1:5556

docker-jmx-prometheus-exporter's Issues

Units on metrics?

Hi there!

First thanks a lot for the exporter. I have a question. Is there somewhere a piece of documentation which maps the /metrics to the provided units?

E.g. does java_lang_MemoryPool_Usage_used deliver bytes or kilobytes?

Or can you update the Readme.md to point to a sample grafana dashboard, where everything is set right to start over with it?

THX in advance
Stefan

Switch to openjdk

Oracle JDK requires commercial license, so we should switch to openjdk as per the guidance from the current anapsix/alpine-java base image

https://hub.docker.com/r/anapsix/alpine-java/

Documentation not up to date on docker-hub

Hi,
I would like to notice you that your documentation on dockerhub : https://hub.docker.com/r/sscaling/jmx-prometheus-exporter/ is not up to date with the github documentation ( still refering the image as sscaling/jmx-exporter instead of sscaling/jmx-prometheus-exporter )
Regards.

Broken in latest

JMX is connecting, but the metrics are not spitting any output. Switching back to 0.3.1 resolves this issue. There's no log output to stdout, so I can't give much more info here.

Some features added to your original skeleton

Hello,

FYI, I have created a dockerized version of jmx-exporter starting from your original skeleton.
https://github.com/eedugon/jmx-exporter-dockerized

I have added some features like dynamic configuration from environment variables provided when creating the docker.

Regards!
Eduardo

Prometheus Jmx Exporter latest docker image has Common Vulnerabilities and Exposures

Following are the details of the critical CVE that has been flagged by our scanner

Image , ID
sscaling/jmx-prometheus-exporter:latest , 0b3f298035c96bdd

CVE , Package , Version ,Severity,Status ,CVSS
CVE-2019-14697 , musl , 1.1.19-r10 , critical , fixed in 1.1.19-r11 , 9.8
CVE-2019-2699 , java , 1.8.0_202 , critical , , 9
CVE-2019-2697 , java , 1.8.0_202 , high , , 8.1
CVE-2019-2698 , java , 1.8.0_202 , high , , 8.1
CVE-2019-2602 , java , 1.8.0_202 , high , , 7.5
CVE-2019-15847 , gcc , 6.4.0-r9 , high , , 7.5
CVE-2018-12886 , gcc , 6.4.0-r9 , high , fixed in 8.3.0-r0 , 8.1
CVE-2018-20679 , busybox , 1.28.4-r3 , high , fixed in 1.30.1-r2 , 7.5
CVE-2019-5747 , busybox , 1.28.4-r3 , high , fixed in 1.30.1-r2 , 7.5
CVE-2019-5482 , curl , 7.61.1-r2 , high , fixed in 7.61.1-r3 , 7.5
CVE-2019-5481 , curl , 7.61.1-r2 high , fixed in 7.61.1-r3 , 7.5