Feature Overview (aka. Goal Summary)  

Include a TCP traffic RTT latency metric in our flow-based observability.
 

Goals (aka. expected user outcomes)

Provide insight to latency-sensitive applications.
 

Requirements (aka. Acceptance Criteria):

•  Loki cannot be a requirement for the latency metric. There must still be a mechanism to query RTT when Loki is not being used. Without Loki, it should still be viewable from Prometheus.

Use Cases (Optional):

Questions to Answer (Optional):

Out of Scope

• Initially, we will focus on TCP only, with roadmap plans to measure latency for all agent-supported protocols (TCP/UDP/SCTP/ICMPv4/ICMPv6). 

Background

Customer Considerations

Documentation Considerations

Interoperability Considerations