feat(infra): observability backplane — Tempo + Prometheus + Grafana (S-16a, closes #122) #125

Merged

not merged 4 commits from feat/122-observability-backplane into main 2026-07-23 12:26:23 +00:00

Conversation 0 Commits 4 Files Changed 13

+263 -3

4 Commits

Author	SHA1	Message	Date
not	56a5fe014c	fix(infra): run the observability check inside the compose network (refs #122) ... The stack runs as sibling containers on the CI runner's host daemon, so its published ports aren't on the runner job's localhost — verify-observability hit localhost:3000 and failed. Reach Grafana by container IP through a throwaway curl container on the stack network, mirroring run-bff-check.sh et al. Parses with grep (the curl image has no python). refs #122	2026-07-23 13:55:03 +02:00
not	43c6c8e0d2	docs: ADR-0023 + demo note + backlog sync for the observability backplane (refs #122) ... ADR-0023 records the Grafana-native stack (Tempo, no collector, config baked into images); demo-script and BACKLOG describe S-16a and the S-16 split. refs #122	2026-07-23 12:48:02 +02:00
not	f29a8598fd	feat(infra): Tempo + Prometheus + Grafana observability backplane in compose (refs #122) ... Three built images on the cg network with config baked in: Tempo (OTLP ingest, ports 4317/4318), Prometheus, and Grafana with both datasources auto-provisioned. Config lives in infra/observability/. The verify-observability step runs early in CI verify-stack; the three containers are added to the failure log-dump list. Tempo's Grafana plugin has no datasource /health method, so the check proves Tempo reachability through Grafana's datasource proxy instead. refs #122	2026-07-23 12:46:54 +02:00
not	f12e9ce746	test(infra): verify-observability asserts Grafana + Tempo/Prometheus datasources are live (refs #122) ... Polls Grafana's datasource-health endpoints so it proves Grafana can actually reach Tempo and Prometheus over the cg network, not merely that the containers started. Fails until the backplane is wired (next commit). refs #122	2026-07-23 12:41:58 +02:00