Agenda
- Bringing back master to "normal" Krzysztof Opasiak
- RC0 Sylvain Desbureaux Krzysztof Opasiak
- review sorting:
- goal: merge as most as possible of the 2 first categories
- for third category, it'll be done in a best effort after the 2 first categories are cleared
- if needed for RC0 (bugfix for component for example), drop an email to Sylvain Desbureaux and Krzysztof Opasiak and they'll move to "prioritized merge list"
- next step: create two new tags: Istanbul and Honolulu
- goal: RC0 to be approved Thursday 25th (so cleared slate for that date)
- Prometheus support: Lukasz Grech
- Allow user to enable monitoring via Prometheus in OOM (via some flag - to be defined)
- When enabled generic ServiceMonitor would be deployed during ONAP installation
- ServiceMonitor would match services with specific label (i.e. metrics: spring-metrics)
- ServiceMonitor would expect predefined URLs to be exposed (i.e. /actuator/prometheus - this is default Spring end-point for Prometheus metrics)
- See attached yaml for ServiceMonitor proposal
- ServiceMonitor would be generic in a sense that it would match all services that are properly labelled. So we would need only 1 ServiceMonitor instance to monitor all services that expose metrics via predefined set of URLs.
- We want to use VES and HV-VES as example use cases, but anyway most of the work will be in OOM.
- OOM team is OK with that as long as metrics are only exposed in the cluster
- Documentation: Sylwia Jakubek Sylvain Desbureaux
- prometheus documentation :
- proposition to remove part exposure ports
- release notes to update
- supported version to update
- bumping version will be done at RC0
{"serverDuration": 400, "requestCorrelationId": "1f1ff0386a728b70"}