Java 17 K8s

java17-in-k8s.sh

# Check java settings within a pod
kubectl exec -it <pod> bash

java -XshowSettings:system -version

# Useful docs at https://developers.redhat.com/articles/2022/04/19/java-17-whats-new-openjdks-container-awareness#tuning_defaults_for_containers

# Setup options via JAVA_TOOL_OPTIONS, https://circleci.com/docs/java-oom/#javatooloptions
# Specify -XX:+UseZGC garbage collector for service with high memory requirements. https://www.linkedin.com/pulse/jdk-17-g1gc-vs-zgc-usage-core-exchange-application-performance-raza
- name: JAVA_TOOL_OPTIONS
  value: -Dlogging.level.root=INFO -XX:+UseZGC -XX:InitialRAMPercentage=40.0 -XX:MaxRAMPercentage=60.0 -XX:ActiveProcessorCount=4
++
resources:
  limits:
    memory: "8Gi"
  requests:
    memory: "2Gi"
    
# Docker images
eclipse-temurin:17-jre-focal
eclipse-temurin:17-jre-alpine
maven:3.8.6-eclipse-temurin-17-focal

# Enable VisualVM 
1. Add environment variable to the k8s deployment. (append the parameters in case the environment variable already exists)
- name: JAVA_TOOL_OPTIONS
              value: "-Dcom.sun.management.jmxremote \
                      -Dcom.sun.management.jmxremote.authenticate=false \
                      -Dcom.sun.management.jmxremote.ssl=false \
                      -Dcom.sun.management.jmxremote.local.only=false \
                      -Dcom.sun.management.jmxremote.port=9010 \
                      -Dcom.sun.management.jmxremote.rmi.port=9010 \
                      -Djava.rmi.server.hostname=127.0.0.1"
2. Add the port in the k8s deployment: 
ports:
- containerPort: 9010
  protocol: TCP
  
3. Do port-forward to the pod 
kubectl port-forward <pod-name> 9010:9010
 
4. Open VisualVM -> Add JMX Connection
   Specify 127.0.0.1:9010 and save it
5. Double click on app.jar