Prometheus Feature and Bug-Fixes

[DanSalt]

• [bugfix] switched message-load-balancing mode to ‘ON_DEMAND’ to enable a fully-symmetric cluster with shared messaging
• [new] turned on the JMX exporter for metrics
• [new] added Prometheus ServiceMonitor to scrape JMX endpoints
• [bugfix] changed charts to use a PV per node, rather than shared, as shared was causing persistence clashes
• [bugfix] changed DNS toleration of unready endpoints to ‘false’ to prevent slave nodes from participating in DNS
• [bugfix] added PodManagementPolicy as ‘Parallel’ to the slave StatefulSet to enable slave nodes to deploy even though they do not reach ‘ready’ state
• [new] added flag to values.yaml to state where Prometheus is deployed (as ServiceMonitor must be deployed in same namespace)

[vromero]

Thank you! this is fantastic. I have left some comments I'm looking forward to read your responses.

[vromero]

Thank you again!

[DanSalt]

No problem at all - glad to be able to contribute. Your chart helped us get up-and-running super quick with Artemis! If we make any other tweaks as we head into production I'll be sure to let you know.