The talk discusses how to make observability actionable with Prometheus and Keptn, using Service Level Objectives (SLOs) and automation to keep cloud native systems reliable and resilient.
- Observability is necessary for making informed decisions about complex cloud native systems
- Prometheus and other frameworks make it easier to obtain observability data
- Service Level Objectives (SLOs) can be used to define the desired state of the system
- Automation can be used to keep the system in the desired state based on observability data
- Keptn is an observability-driven orchestration engine for cloud native apps that uses SLOs and automation to keep systems reliable and resilient
The speaker mentions a post from a performance engineer at Facebook who saw the potential of Keptn to bring SRE core principles to the cloud native community by using observability data to drive actions that keep the system reliable.