Monitoring RHODS components is essential for:

• Easier support from MTSRE

• Visibility on SLA/SLO for RHODS

• Visibility for RHODS developer for resolving issues
• Hybrid SRE https://source.redhat.com/groups/public/sdsea/hybridsre

Keeping those goals in mind, we mostly need metrics about availability and performance of each RHODS components.

Requirements:

• the monitoring is adaptable to the addon flow and the self-managed flow
• we can federate metrics across RHODS installation by storing them in a remote store (RHOBS or other), to give access to metrics without access to customer clusters
• use standard components, in particular for all the RHODS operator based on operator-sdk, which already provides standard metrics. (https://book.kubebuilder.io/reference/metrics-reference.html)