Hacktoberfest 2026:メンテナが10月に向けて印を付けた、オープンで初心者向けの issue。 Hacktoberfest の issue を見る

Update /ready and /health schemas for v2.10

オープン
#659 コメント 1 件 リアクション 0 件 担当者 2 名 GitHub で見る

@devanbenz がすでに取り組んでいます。

2026年10月9日 から。

  • #661 @jstirnaman による — オープン

評価

この issue はまだ評価されていません。

説明

Summary

Two contract changes are needed for InfluxDB OSS v2.10 and InfluxDB Cloud 2:

  1. The /ready and /health schemas don't describe the v2.10 responses.
  2. The Cloud Routes schema is missing two keys that Cloud returns.

I tested influxdata/influxdb main-2.x at 310ee146d9 (built locally) and a production InfluxDB Cloud 2 (TSM) org on 2026-10-08.

1. /ready and /health schemas are out of date

Problem

src/common/schemas/Ready.yml and src/common/paths/ready.yml haven't changed since 2021-03.
main-2.x returns fields and states that neither contract describes.
HEALTH_READY.md in main-2.x documents the same behavior.

influxdata/docs-v2#7286 added some of these changes by editing the OSS spec in docs-v2 directly.
The next sync from this repo would overwrite them, so the change belongs here.

Verification

/ready during startup returns 503:

{"status":"starting","started":"2026-10-08T17:54:06.861243-05:00","up":"48.001208ms",
 "uptime":{"value":0.048001208,"unit":"seconds"},
 "checks":[{"name":"shards","status":"fail","message":"waiting for shard enumeration",
            "progress":{"completed":0,"total":0,"unit":"shards"}}, ...]}

/ready after startup returns 200 with status, started, up, and uptime, and no checks.

/health returns 200 with a failures object in the shards check, plus version and commit:

{"name":"influxdb","status":"pass","message":"healthy",
 "checks":[{"name":"shards","status":"pass","failures":{"count":0,"unit":"shards"}}, ...],
 "version":"dev","commit":"none"}

After a failed startup, /health returns 503 with status: fail, and the failing check's message contains the raw error.
With --startup-error-linger, both endpoints keep serving for the linger window before influxd exits.

With --health-auth-mode=required, the status code stays the same for every caller.
Only a caller with an operator token receives check messages, measure groups, version, and commit.
Every other caller receives check names and statuses only.
I tested with no token, a read-only token, an invalid token, and an operator token.

Suggested solution

Update the common /ready and /health schemas and paths to cover:

  • The 503 response on /ready, with status: starting and only the failing checks in checks.
  • uptime on /ready, as {"value": <seconds>, "unit": "seconds"}.
  • The measure-group objects on checks, such as progress and failures.
  • version and commit on /health.
  • A description note: --health-auth-mode and --hardening-enabled reduce the response body for callers without operator permissions, and never change the status code.

After the change ships, docs-v2 closes influxdata/docs-v2#7286 and picks up the change through a normal sync.

主要言語
Shell
スター
18
フォーク
9
平均マージ
2日 18時間
マージ済み PR(30日)
1

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。

influxdata/openapi のほかの issue

influxdata/openapi の issue をすべて見る

似ている issue

Shell/Bash 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。