Release: unfreeze agent status view + smoothed throughput autoscaler + log pane #165

2026-06-30T20:20:27-04:00

bvandeusen commented

2026-06-30 20:20:27 -04:00

Agent-only. Fixes the operator report that the status tiles (state/active/processed) and Start/Stop freeze while the GPU meters stay live.

Root cause: /status made an inline blocking queue_status call to a curator buried under a 112k backlog → the whole status refresh stalled (/gpu survived because it's a lock-free local read). Compounded by the old util-band autoscaler growing to the 32-worker cap forever (util plateaus ~50% on this IO-bound load, never hit the 70 threshold), piling on load.

/status is now a pure in-memory read — status() is lock-free and the queue snapshot comes from a background poller.
Autoscaler → smoothed, throughput-aware climb that settles: EWMA-smoothed util, grow-while-jobs/s-improves, back off + hold when a grow stops helping. No runaway, no flopping.
GPU util bar shows a smoothed value (agent EWMA util_smooth on /gpu, else client-side) instead of bouncing 0↔99.
act() aborts a slow POST after 8s so buttons can't stick.
Log pane bounded to the viewport (scrolls internally, no off-screen overflow); ring buffer capped at 400 lines.

Ships in the agent image.

🤖 Generated with Claude Code

Agent-only. Fixes the operator report that the status tiles (state/active/processed) and Start/Stop freeze while the GPU meters stay live. **Root cause:** `/status` made an inline blocking `queue_status` call to a curator buried under a 112k backlog → the whole status refresh stalled (`/gpu` survived because it's a lock-free local read). Compounded by the old util-band autoscaler growing to the 32-worker cap forever (util plateaus ~50% on this IO-bound load, never hit the 70 threshold), piling on load. - `/status` is now a pure in-memory read — `status()` is lock-free and the queue snapshot comes from a background poller. - Autoscaler → smoothed, throughput-aware climb that **settles**: EWMA-smoothed util, grow-while-jobs/s-improves, back off + hold when a grow stops helping. No runaway, no flopping. - GPU util bar shows a **smoothed** value (agent EWMA `util_smooth` on `/gpu`, else client-side) instead of bouncing 0↔99. - `act()` aborts a slow POST after 8s so buttons can't stick. - Log pane bounded to the viewport (scrolls internally, no off-screen overflow); ring buffer capped at 400 lines. Ships in the agent image. 🤖 Generated with [Claude Code](https://claude.com/claude-code)

bvandeusen added 1 commit 2026-06-30 20:20:27 -04:00

fix(agent): unfreeze status view + smoothed throughput-aware autoscaler + log pane

CI / lint (push) Successful in 3s

Details

CI / frontend-build (push) Successful in 21s

Details

CI / backend-lint-and-test (push) Successful in 26s

Details

CI / integration (push) Successful in 3m23s

Details

f0f031782d

Operator: the status tiles (state/active/processed) and the Start/Stop buttons
freeze while the GPU meters stay live. Root cause: /status made an INLINE
blocking curator call (queue_status) on every poll, and with curator buried
under a 112k-job backlog that call stalled — freezing the whole status refresh
(the GPU bars survived because /gpu is a lock-free local read). Made worse by the
old util-band autoscaler, which grew workers toward the 32 cap forever because
util plateaus ~50% on this IO-bound load and never hit the 70 grow threshold —
piling load onto curator and the agent process.

- /status is now a pure in-memory read: worker.status() is lock-free, and the
  curator queue snapshot is refreshed by a background poller (never inline).
- Autoscaler replaced with a smoothed, throughput-aware climb that SETTLES:
  samples util every 2s and EWMA-smooths it (raw util swings 0↔99), then every
  ~24s grows by one only while each grow keeps lifting smoothed jobs/s; when a
  grow stops helping it backs off one and holds, re-probing occasionally. No
  runaway, no flopping.
- GPU util bar now shows a smoothed value: the agent's own EWMA (util_smooth,
  exposed on /gpu) when running, else smoothed client-side — so it glides
  instead of bouncing 0↔99.
- act() aborts a slow Start/Stop POST after 8s so the buttons can't stick; the
  now-always-fast /status refresh recovers state regardless.
- Log pane: bound the page to the viewport (height:100vh) so the Logs card
  scrolls INTERNALLY instead of overflowing off-screen; cap the ring buffer at
  400 lines.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Ttrj5P7upUTueSfoJcxEqa

bvandeusen merged commit a28c33281a into main

2026-06-30 20:20:34 -04:00

Sign in to join this conversation.

1 Participants

Notifications

Due Date

No due date set.

Dependencies

No dependencies set.

Reference: bvandeusen/FabledCurator#165