Currently the administration guide of ONAP is poor, it could make sense to plan an update of the documentation in the future.

...

•  Which tools are used: Nagios, Prometheus, Zabbix, Consul, proprietary third party tooling,...?    We actively use Zabbix for monitoring

• What is your feedback on the tools?

           we use Zabbix monitors the CPU, memory, and disk information of k8s nodes mainly by triggering alarm and sending messages by emails or others ways when a montior thereshold is exceed to users who is maintaining the env, and help uses to solve the problems.

• how do you manage Infra versus ONAP versus VNF supervizion?Not considered yet.
• Do you use a ligther tools to check the end points for the ONAP end users?

          Not considered yet.

Orange OpenLab (Morgan Richomme Eric Debeau)

•  Which tools are used: Nagios, Prometheus, Zabbix, Consul, proprietary third party tooling,...?

...

• Do you use a ligther tools to check the end points for the ONAP end users?

ONAP Lab for OVP(Sriram Rupanagunta)

•  Which tools are used: Nagios, Prometheus, Zabbix, Consul, proprietary third party tooling,...?

...

• What is your feedback on the tools?

• how do you manage Infra versus ONAP versus VNF supervizion?

• Do you use a ligther tools to check the end points for the ONAP end users?

Bell Canada production ONAP(Jean-Francois Patenaude)

•  Which tools are used: Nagios, Prometheus, Zabbix, Consul, proprietary third party tooling,...? 
  • We actively use:
    • Zabbix for monitoring
    • Jira Service Desk for ticketing
    • Slack for live notifications
    • OpsGenie for alerting (Page)
      Image Added
  • We also use:
    • Elastic (ELK) for pods logs aggregation and some dashboards
    • Prometheus/Grafana for some dashboards

• What is your feedback on the tools?
  • Zabbix has a steep learning curve but when you know how it works, it is very powerful.  Triggers are easy to configure and they can be advanced if you need. It's even possible to do nice dynamic dashboards like this:
    Image Added
    For info: all the green checkmarks are actual live monitoring. For instance, if an ONAP component gets in problem, the green checkmark Image Added is replaced by a red X mark Image Added.
    These kind of dashboards let you quickly diagnose where the problems are located.
  • We configured Zabbix to only collect metrics related to things it will alert on. It's nearly a 1:1 relationship between what we gather and what we alert on.
  • Prometheus/Grafana is useful to gather lots of metrics and build dashboard. We are not currently alerting using these metrics. When there is a metric we like, we send it in Zabbix instead for the alerting.

• how do you manage Infra versus ONAP versus VNF supervizion?

...