Memory pressure: the console shows a worker's step
I maintainer di solito rispondono entro 1 giorno
Nessuno ha ancora preso questa issue.
Valutazione
- Difficoltà
- 3/5
- Tempo stimato
- 1-2 giorni
- Idoneità per principianti
- 74/100
- Tipo di issue
- Funzionalità
- Chiarezza
- Specificata chiaramente
- Stato di attività
- Attiva
- Stack tecnologico
- typescript
- Ambito
- frontend, observability
Direzione di ricerca
Start with ui/src/model/fleet.ts, worker-page.ts, and stale-keys.ts; the issue identifies each model seam and the corresponding tests. Read the referenced tests and ui/AGENTS.md first, then run the listed Vitest command. Done means the step appears with its word and tone in the Fleet square and Memory card, the event marks workers stale, and the unit tests and pnpm lint pass.
Scritto dal modello di indicizzazione a partire dal testo della issue.
Descrizione
Part of #1355.
Scope
An operator sees a worker's memory-pressure step in the console. The Fleet row's square names it, and the worker page's Memory card has a row for it, on a gateway and on a standalone worker alike. A memory.pressure-changed event refreshes what the console shows.
In short
flowchart LR
S[status.get capacity.memoryPressure] --> F[Fleet row square: findings *new word*]
S --> W[worker page Memory card: Paused row *new*]
E[memory.pressure-changed] -->|*new* marks workers stale| S
| Step | Fleet square word and colour | Worker page Memory card row |
|---|---|---|
new-devices-paused |
new devices paused, red |
Paused: new devices (red) |
warming-paused |
warming paused, amber |
Paused: warming (amber) |
none, or no field |
nothing new | no row |
| worker not connected | nothing new: disconnected already wins, as today (ui/src/model/fleet.ts:31) |
no Memory card, as today |
The square still shows the worst finding, and each colour keeps its word. Of two findings with the same colour the square shows the first (F7), so the step's finding goes before the pressure and figure findings: at warming-paused with memory pressure warn the square reads warming paused, and at new-devices-paused with memory pressure critical it reads new devices paused.
If this goes wrong, the console shows a worker as calm while it refuses new devices, or the other way round.
Technical spec
Facts
Commit 6674ad0c.
| # | Fact | Source |
|---|---|---|
| F1 | The Fleet row's findings come from loadFindings: none for a worker not connected; memory pressure warn/critical with tone warn/error |
ui/src/model/fleet.ts:45-65 |
| F2 | The worker page's Memory card takes its tone from the pressure and lists a Pressure detail row |
ui/src/model/worker-page.ts:207-211, ui/src/model/worker-page.ts:233-245 |
| F3 | An event's stale keys come from staleKeys; capacity.changed marks the workers key; no test file covers it |
ui/src/model/stale-keys.ts:18-31, rg -l "staleKeys" ui/src e2e printed only stale-keys.ts and ui/src/state/queries.ts |
| F4 | The Fleet tests cover the square's findings, pressure included | ui/src/model/fleet.test.ts:84 |
| F5 | The console page documents the square's colours and words | website/src/content/docs/docs/worker/console.mdx:104-108 |
| F6 | Colour is never the only signal; orange is never a state | ui/AGENTS.md (Look) |
| F7 | workerHealth keeps the first finding of the worst tone; loadFindings pushes the pressure finding last |
ui/src/model/fleet.ts:19-22, ui/src/model/fleet.ts:46-65 |
Modules touched
Searches: loadFindings, pressureStatus, figureCards, staleKeys, capacity.changed in ui/src.
flowchart LR
SC[screens fleet, worker page] --> ST[state hooks]
ST --> M1[model fleet.ts *]
ST --> M2[model worker-page.ts *]
Q[state queries.ts] --> M3[model stale-keys.ts *]
ui/src/model/fleet.ts: a finding fromworker.capacity?.memoryPressure, placed before the pressure and figure findings (F7).ui/src/model/worker-page.ts: thePauseddetail row on the Memory card.ui/src/model/stale-keys.ts:memory.pressure-changedmarks the workers key; a newui/src/model/stale-keys.test.ts.- Docs: website
worker/console.mdx(the square table),fleet-console.mdxif it lists the words.
Contract and event changes
none: the console reads capacity.memoryPressure (#1361) and memory.pressure-changed (#1361) as types only.
Other code on the same state
Searches: readers of worker.capacity and worker.load.memory in ui/src/model (fleet.ts, worker-page.ts, overview.ts).
overview.tsreads capacity figures for counts only, not the step; unchanged.- The pressure finding (F1) and the step finding can both show; the square takes the worse tone, as today.
- On a gateway, the console sees a new step only after the gateway refreshes that worker's view; until #1364 adds the event to the gateway's refresh list, that is its 30 s tick. Accepted: no Depends on #1364.
- #1145 and #1156 (other features' drafts) edit the
staleKeysswitch andworker-page.ts; whichever lands second rebases. No Depends on: they are not ready.
Failure modes
none: the change crosses no boundary and writes no state.
Rules in play
- ui manual: "Colour is never the only signal"; every new tone carries its word.
- ui manual: "
src/model/has no React and no network, so every rule is a plain unit test." - ui manual: "An event marks keys stale through
src/model/stale-keys.ts."
Tests
The seam is the model builders, modelled on ui/src/model/fleet.test.ts and ui/src/model/worker-page.test.ts.
- An operator's console shows a worker's step on its card and page:
new-devices-pausederror tone with words;warming-pausedwarn tone;warming-pausedbesidememory pressure warnreadswarming paused;new-devices-pausedbesidememory pressure criticalreadsnew devices paused; none nothing. - A
memory.pressure-changedevent marks the console's workers stale: the event marks workers; an unrelated event does not.
Done when
pnpm exec vitest run --project unit ui/src/model/fleet.test.ts ui/src/model/worker-page.test.ts ui/src/model/stale-keys.test.tspasses.pnpm lintpasses.
Out of scope
- A new screen or a chart of the step over time.
Depends on
- #1361
Written by an agent.
- Lingua principale
- TypeScript
- Stelle
- 535
- Fork
- 21
- Merge medio
- 9h 10m
- PR unite (30g)
- 454
Preparare l'ambiente
Questo progetto non fornisce container di sviluppo, Dockerfile né guida per i contributori, quindi l'ambiente è a tuo carico: parti dal suo README e consulta la nostra guida al primo contributo per i passaggi generali.
Come iniziare
- Leggi tutta la issue e poi la guida ai contributi del progetto.
- Commenta sulla issue per dire che te ne occupi tu — evita che due persone facciano lo stesso lavoro.
- Fai un fork del repository e lavora su un branch.
- Apri una pull request che faccia riferimento al numero della issue.
Altre issue di callstackincubator/simlock
-
os:windows request:new
Difficoltà 2/5 1-3 ore Idoneità per principianti 70/100
callstackincubator/simlock#1071 ·
I maintainer di solito rispondono entro 1 giorno
-
A learned kind keeps one ready copy while its copies are leasedForse già presa @V3RON l’ha presa oggi. Apertatask:blocked
Difficoltà 4/5 3-5 giorni Idoneità per principianti 20/100
callstackincubator/simlock#1381 · 1 commento ·
I maintainer di solito rispondono entro 1 giorno
-
Warm pool keeps a spare learned copy while its copies are heldForse già presa Una pull request collegata a questa issue è aperta o già unita. Apertafeature:planned
Difficoltà 4/5 3-5 giorni Idoneità per principianti 35/100
callstackincubator/simlock#1374 · 2 commenti ·
I maintainer di solito rispondono entro 1 giorno
-
request:new
Difficoltà 3/5 1-2 giorni Idoneità per principianti 55/100
callstackincubator/simlock#1372 ·
I maintainer di solito rispondono entro 1 giorno
-
bug:new
Difficoltà 4/5 1-2 giorni Idoneità per principianti 55/100
callstackincubator/simlock#1371 · 2 commenti ·
I maintainer di solito rispondono entro 1 giorno
Tutte le issue di callstackincubator/simlock
Issue simili
-
community first-timers-only good first issue hacktoberfest help wanted low hanging fruit up-for-grabs
Difficoltà 1/5 Meno di un'ora Idoneità per principianti 83/100
lingdojo/kana-dojo#33701 · 1 commento · 5 reazioni ·
I maintainer di solito rispondono entro 1 giorno
-
tech-debt
Difficoltà 2/5 1-3 ore Idoneità per principianti 68/100
rookslog/zlibrary-mcp#408 ·
I maintainer di solito rispondono entro 1 giorno
-
mp4box.js — 20-byte `sgpd` box with large `entry_count` causes unbounded loop and heap OOM abortAperta
Difficoltà 2/5 1-3 ore Idoneità per principianti 77/100
I maintainer di solito rispondono entro 1 giorno
-
Difficoltà 2/5 1-3 ore Idoneità per principianti 78/100
I maintainer di solito rispondono entro 1 giorno
-
status: needs triage
Difficoltà 2/5 1-3 ore Idoneità per principianti 64/100
mastra-ai/mastra#29641 · 1 commento ·
I maintainer di solito rispondono entro 1 giorno