Files
orca/config/scripts/run-multi-workspace-typing-bench.mjs
T
Jinwoo-HandOrca d0b067e7f0 perf(daemon): keep-tail thin hidden panes' stream so agent floods never bury typing (STA multi-workspace lag)
Hidden panes are exempt from pendingData flow control (main gate-drops
their bytes after ingestion), so N background agents ran unbounded ahead
on the one shared daemon->main stream socket (measured 192MB user-space
backlog) and visible-pane echo waited FIFO behind it — typing appeared
seconds late whenever several agents burst on a loaded machine
(8x512KB/s + 12 CPU spinners: p50 293ms fix-off; 12x1MB/s: 6.1s).

Mechanism (replaces producer pacing — no reveal catch-up, ever):
- Shallow socket write gate (128KB) + per-session fairness bypass bounds
  echo latency by construction; kernel-flush refill sentinel keeps held
  bulk draining at full speed (drain-only refill capped at ~8MB/s).
- Backgrounded sessions' queued output is keep-tail dropped (newest
  512KB kept, in-order dataGap replaces the middle); a ~2MB GLOBAL
  budget shrinks per-session keep-tails (floor 64KB) so a worktree
  switch never waits behind the aggregate. Reply-eliciting query bytes
  (DSR/DA/OSC probes) are salvaged from dropped spans.
- Notifications are structurally lossless: the daemon runs the same
  shared scanners main uses (bell/OSC 133/pr-link/2031) over every byte
  BEFORE drop decisions and relays facts in byte order; ordered
  background markers hand scan authority back and forth, seeded with the
  emulator's partial escape tail so a sequence split across the handoff
  neither phantom-fires nor goes missing. Titles/agent-status stay
  main-side (kept-tail convergent).
- Main: background = hidden AND no remote view subscriber (a live
  mobile/web view is never thinned); on dataGap main resets cross-chunk
  parse carries, drops the headless mobile mirror, and reuses the
  hidden-drop model-restore marker.

Wire: three new stream events, tolerated within protocol v19 (old mains
ignore unknown events; old daemons never see the trigger). Kill
switches: ORCA_DAEMON_BACKGROUND_STREAM_DROP=0,
ORCA_DAEMON_SHALLOW_SOCKET_GATE=0.

A/B (pnpm bench:multi-workspace-typing): 8x512KB/s + 12 CPU workers
p50 293ms/p90 647ms -> 15/21ms (= baseline); 12x1MB/s 6,146ms -> 20ms;
light loads unchanged; zero missing echoes. Latin hidden-restore e2e
green (probe-verified aggregate-drain root cause). New deterministic
repro harness: tests/e2e/terminal-multi-workspace-typing-latency.spec.ts
+ CPU pressure workers.

Co-authored-by: Orca <help@stably.ai>
2026-07-07 15:41:05 -04:00

64 lines
1.6 KiB
JavaScript

/**
* Entry point for the multi-workspace typing-latency bench
* (tests/e2e/terminal-multi-workspace-typing-latency.spec.ts).
*
* Usage:
* pnpm bench:multi-workspace-typing [-- --panes 8 --rate-kbps 512 \
* --keys 48 --cadence-ms 250 --cpu-workers 4 --label before-fix]
*
* Results land in tools/benchmarks/results/multi-workspace-typing-*.json.
* Run once per build/config with distinct --label values, then diff the
* totalMs/inputHalfMs/echoHalfMs percentiles.
*/
import { spawn } from 'node:child_process'
const npxCommand = process.platform === 'win32' ? 'npx.cmd' : 'npx'
const knobByFlag = {
'--panes': 'ORCA_TYPING_BENCH_LOAD_PANES',
'--rate-kbps': 'ORCA_TYPING_BENCH_RATE_KBPS',
'--keys': 'ORCA_TYPING_BENCH_KEYS',
'--cadence-ms': 'ORCA_TYPING_BENCH_KEY_CADENCE_MS',
'--cpu-workers': 'ORCA_TYPING_BENCH_CPU_WORKERS',
'--label': 'ORCA_TYPING_BENCH_LABEL'
}
const env = { ...process.env, ORCA_TYPING_BENCH: '1' }
const passthroughArgs = []
const argv = process.argv.slice(2)
for (let i = 0; i < argv.length; i++) {
if (argv[i] === '--') {
continue
}
const knob = knobByFlag[argv[i]]
if (knob) {
env[knob] = argv[++i]
} else {
passthroughArgs.push(argv[i])
}
}
const child = spawn(
npxCommand,
[
'playwright',
'test',
'tests/e2e/terminal-multi-workspace-typing-latency.spec.ts',
'--config',
'tests/playwright.config.ts',
'--project',
'electron-headless',
'--workers=1',
...passthroughArgs
],
{ stdio: 'inherit', env }
)
child.on('exit', (code, signal) => {
if (signal) {
process.kill(process.pid, signal)
return
}
process.exit(code ?? 1)
})