ISD-1729 Add RAM percentage JVM option and documentations (#134)

docs/how-to/configure-jenkins-memory-usage.md

@@ -0,0 +1,13 @@
+# How to control heap memory of the jenkins-k8s-operator charm
+The [jenkins-k8s-operator](https://github.com/canonical/jenkins-k8s-operator) charm uses [juju constraints](https://juju.is/docs/juju/constraint) to limit the amount of memory a charm can use. To deploy the charm with constraints, use the `--constraints "<key>=<value>"` option when running `juju deploy`:
+```bash
+juju deploy jenkins-k8s --channel=latest/edge --constraints "mem=2048M"
+```
+To change this value after deployment, use the `set-constraints` command.
+```bash
+juju set-constraints jenkins-k8s "mem=4096M"
+```
+Other types of constraints (like cores, disk, etc.) can also be applied. Note that this value affects the shared maximum memory between the `charm` container and `jenkins` container.
+
+# Considerations when applying memory constraints
+Constraints set this way directly influence the amount of heap memory available to the JVM, with a ratio `JVM heap / Container Memory limit` of 0.5. For example, a `jenkins-k8s-operator` charm deployed with `--constraints "mem=1024M"` would set a maximum heap memory size of 512Mb. Too little heap memory can result in the controller getting restarted due to Out-of-memory(OOM) error. Make sure to adapt the memory constraints based on your workload.

src/pebble.py

@@ -78,6 +78,7 @@ def _get_pebble_layer(jenkins_instance: jenkins.Jenkins) -> ops.pebble.Layer:
                 "summary": "jenkins",
                 "command": f"java -D{jenkins.SYSTEM_PROPERTY_HEADLESS} "
                 f"-D{jenkins.SYSTEM_PROPERTY_LOGGING} "
+                "-XX:MaxRAMPercentage=50.0 -XX:InitialRAMPercentage=50.0 "
                 f"-jar {jenkins.EXECUTABLES_PATH}/jenkins.war "
                 f"--prefix={env_dict['JENKINS_PREFIX']}",
                 "startup": "enabled",