Investigate serverError cluster on /practicas (matias.garciablanco@odontologia.uba.ar, 2026-06-18 20:28, concentrated single-session): 5x /practicas render (digest 163624117) + 2x /practicas/[id] action (1630870042) + 1x /practicas/[id] render + 1x / action. GENUINE 500s, real user, one tight window. Also note older /administrar/alumnos/[id] render 500 digest 3278366549 (8 hits 2026-06-10/11, stopped) - look up that digest too. Find root cause.

Sub-tickets

Questions

Activity

• wi-cli-venus created · 2026-06-22
• wi-cli-venus assigned · 6w ago
  coder-pluto-cc / coder
• wi-cli-venus commented · 6w ago
  Queued behind 188 (same coder). Investigate-first: /practicas 500 cluster digests 163624117 + 1630870042 + older alumnos/[id] 3278366549.
• wi-cli-venus completed · 6w ago
  Already-resolved (same appEvents-stale pattern as 188). Coder root-caused: /practicas 500 cluster = 11x pg-pool acquire-timeout (max:2 pool drained by ~2.4s filter-option queries, PLUTO-93/471 mechanism) — matias's 06-18 20:28 session hammered /practicas + /practicas/[id] and starved concurrent renders on pool-acquire. PLUTO-472's 24h option cache (3ac9222) EXPLICITLY covers /practicas comisiones too → already shipped. Plus 1x benign Next16 RSC router-state-header transient (not actionable). DECISIVE: count(/practicas level=error) since 472-deploy 07-02 18:43 = ZERO; latest of any = 06-24, all pre-472; prod v1.97.34. Older digest 3278366549 (/administrar/alumnos/[id]) = 'Sin permisos' gate-correct 403s logged at error level, ceased 06-11 — NOT a crash, telemetry noise → PLUTO-482. No code change. Caveat: /practicas/[id] detail path isn't option-cached (its 06-18 timeouts were contention collateral); belt-suspenders durability lever = pool headroom → noted on PLUTO-146.