Console: manage devices, the warm pool and settings from the console

オープン
#1,130 コメント 4 件 リアクション 0 件 担当者 0 名 GitHub で見る

メンテナーはふだん 1 日以内に返信

まだ誰も着手していません。

評価

難易度
5/5
見積もり時間
1週間以上
初心者へのやさしさ
35/100
issue の種類
機能追加
明瞭さ
おおむね明確
活発さ
活発
技術スタック
json, typescript
領域
api, backend, frontend

調査の方向性

Read ADR 0037, ADR 0017, ADR 0038, and docs/internal/IDEAS.md to understand the console's HTTP-only action model and the deferred design decisions. The issue identifies existing POST /v1/devices and DELETE /v1/devices/{id} routes, but items 3–5 need design decisions first; begin by clarifying which scope is approved. Done means implementing only the agreed console actions and verifying their behavior, including confirmation before irreversible actions and preventing deletion of leased devices.

索引モデルが issue の本文から書いたものです。

説明

request:new
The problem

An operator who runs a worker (or a fleet behind a gateway) has to leave the console for the CLI to manage the devices themselves. The console shows every device and can already act on leases, requests and workers (ADR 0037), but it cannot:

  • enrol a USB phone, or remove an enrolled one;
  • delete an idle simulator or emulator it no longer wants;
  • add a virtual device of a chosen model and OS, so it exists before the first agent asks for it;
  • warm a device up (boot it and keep it ready) or warm it down (shut it down), or choose which devices the warm pool keeps running;
  • change any setting. Warm-pool targets, idle timeouts, emulator flags, stream and agent-device options all live in config.json and change only through simlock config set or by editing the file. There is no settings page and no HTTP route to read or change config.
Who runs into this

A developer running one worker for their own coding agents, and an operator running a fleet. Both watch the console all day. When they want a different set of devices running (for example, "keep one Pixel 8 and one iPhone 17 warm for my agents, shut the rest down"), they have to switch to a terminal, find the right command or config key, and often restart the daemon for a warm-pool change to apply.

What you would like to happen

From the console, on a worker (and through the gateway for a fleet):

  1. Physical devices: enrol a connected phone and remove an enrolled one. The routes exist already (POST /v1/devices, DELETE /v1/devices/{id}).
  2. Delete an idle virtual device, behind a confirm, never one that is leased.
  3. Add a virtual device of a model and OS from the catalog.
  4. Warm up / warm down: boot a device and keep it ready, or shut it down; pick which devices or kinds the warm pool keeps running, without a restart.
  5. Settings: see the daemon's settings and change the ones that are safe to change at runtime.

Items 1 and 2 seem to fit the console's current rules (act only through HTTP routes, one action list per item, confirm what cannot be undone). Items 3 to 5 touch ideas the docs defer today (docs/internal/IDEAS.md: "changing targets without a restart", "fleet-level targets on a gateway") and need a design decision first.

Anything else
  • Related: ADR 0037 (the console acts through the HTTP routes), #769 (release, cancel, drain, undrain and remove from the console), ADR 0017 and ADR 0038 (warm pool).
  • Workaround today: simlock device add|remove, simlock config set warmPool.targets … plus a daemon restart, simlock cleanup / nuke.

Written by an agent.

主要言語
TypeScript
スター
705
フォーク
40
平均マージ
12時間 9分
マージ済み PR(30日)
376

環境構築

このプロジェクトには開発コンテナ、Dockerfile、コントリビューションガイドがありません。まず README を読み、一般的な手順ははじめてのコントリビューションガイドを参照してください。

はじめの一歩

  1. issue を最後まで読み、次にプロジェクトのコントリビューションガイドを読みます。
  2. 着手することを issue にコメントします — 二人が同じ作業をするのを防げます。
  3. リポジトリをフォークし、ブランチを切って変更します。
  4. issue 番号を参照したプルリクエストを送ります。

callstackincubator/simlock のほかの issue

callstackincubator/simlock の issue をすべて見る

似ている issue

TypeScript の issue をもっと見る

新しい issue をメールで受け取る

初心者向けの GitHub issue を短くまとめたダイジェスト。