Artemis Monitoring With Grafana

1.  Install Artemis
Download artemis from  https://activemq.apache.org/artemis/download.html

 Unzip artemis to a location
Lets call the location where Artemis is installed as ARTEMIS_HOME

2.  Create a messaging broker
 
Choose a new location to store broker. Lets call this location as "ARTEMIS_BROKERS"
 
cd to "ARTEMIS_BROKERS"

Run command  :     ./${ARTEMIS_HOME}/bin/artemis create "firstbroker"

Command will result in creation of broker with name "firstbroker" inside        "ARTEMIS_BROKERS" location 

Test the broker  by 

Goto to directory "${ARTEMIS_BROKERS}/firstbroker/bin"

Run command   ./artemis run  (on windows  artemis.cmd run)

If the everything went well  you should be able to access artemis  console  at  http://localhost:8161/console

3. Next step is to export Jmx as metrics using Prometheous jmx exporter javaagent

Download the javaagent jar jmx_prometheus_javaagent-0.3.2-SNAPSHOT.jar from  https://github.com/prometheus/jmx_exporter

Move jmx_prometheus_javaagent-0.3.2-SNAPSHOT.jar to ${ARTEMIS_HOME}/lib

You can read more about Prometheus jmx export here  https://github.com/prometheus/jmx_exporter

Download artemis-2.yml from https://github.com/prometheus/jmx_exporter/tree/master/example_configs

Move the file to ${ARTEMIS_BROKERS}/firstbroker/etc"

Stop the broker if already running

Edit command in artemis.sh file  where the  "exec"  makes call to java to launch  org.apache.activemq.artemis.boot.Artemis

Add the entry for javaagent and modified command will look like

exec "$JAVACMD" $JAVA_ARGS $ARTEMIS_CLUSTER_PROPS \
-javagent:"${ARTEMIS_HOME}/lib/jmx_prometheus_javaagent-0.3.2-SNAPSHOT.jar"="8080:${ARTEMIS_BROKERS}/firstbroker/etc/artemis-2.yml"
-classpath "$CLASSPATH" \
-Dartemis.home="$ARTEMIS_HOME" \
-Djava.library.path="$ARTEMIS_HOME/bin/lib/linux-$(uname -m)" \
$DEBUG_ARGS \
org.apache.activemq.artemis.boot.Artemis "$@"

Note :  For windows mention  relative paths for -javaagent. Giving full path  ( due to colon bein separater in javaagent options) will result in error.

Restart the broker

./artemis run  (on windows  artemis.cmd run)

Call http://localhost:8080/metrics  from browsers.

Details of Jvm is coming on the screen.

4. Enable Local Monitoring and Management to expose MBean showing broker details

Next step is to expose MBean showing broker details

To allow the JMX client access to a local Java VM, you had to set the  system property com.sun.management.jmxremote to true for the for the Artemis java process.

Update command to add -Dcom.sun.management.jmxremote=true

Final command looks like

exec "$JAVACMD" $JAVA_ARGS $ARTEMIS_CLUSTER_PROPS \
-javagent:${ARTEMIS_HOME}/lib/jmx_prometheus_javaagent-0.3.2-SNAPSHOT.jar"="8080:${ARTEMIS_BROKERS}/firstbroker/etc/artemis-2.yml
-classpath "$CLASSPATH" \
-Dartemis.home="$ARTEMIS_HOME" \
-Dcom.sun.management.jmxremote=true \
-Djava.library.path="$ARTEMIS_HOME/bin/lib/linux-$(uname -m)" \
$DEBUG_ARGS \
org.apache.activemq.artemis.boot.Artemis "$@"


Restart the broker

Call http://localhost:8080/metrics  from browsers.

Now the broker details is coming in response.


5. Install and Run Prometheus server
   
 Download Prometheus server  from https://prometheus.io/download/ and untar in a location. Lets call Prometheus server location to ${PROMETHEUS_HOME}

Update prometheus.yml  file located inside  ${PROMETHEUS_HOME}

Updated scrape_configs section looks like. We have update the targets to point to the metrics exposed by Prometheus jmx exporter agent

   scrape_configs:
  # The job name is added as a label `job=<job_name>` to any timeseries scraped from this config.
  - job_name: 'prometheus'

# metrics_path defaults to '/metrics'
# scheme defaults to 'http'.

static_configs:
- targets: ['localhost:8080']


Start the Prometheus server
Browse to  http://localhost:9090

Metrics are available on graph page

6.  Create Grafana dashboard for Artemis
 
 Download Grafana from https://grafana.com/grafana/download  based on the platform and unzip to a location .

Lets call the location as GRAFANA_HOME

Start the grafana server.

Launch the grafana console. http://localhost:3000

Create the new datasource named  DS_PROMETHEUS   of type premetheus with url pointing to Prometheus server(localhost:9090  in this case)

Create Artemis dashboard by importing  file from https://grafana.com/api/dashboards/9087/revisions/1/download

Change the instance variable to localhost:8080

 All done !!