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

UC secrets: supported access to resolved non-secret variables and redaction compatibility

オープン
#7,099 コメント 0 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
go
領域
cli

調査の方向性

Start with bundle/direct/dresources/redact_config.go at v1.17.0, which the issue identifies as the location of the string-only redaction check. Review the validation JSON output and bundle variable resolution path to determine whether a supported way to expose selected non-secret resolved values exists. Done means maintainers clarify the compatibility contract and document an existing workflow or define a supported interface with sensitive value handling.

索引モデルが issue の本文から書いたものです。

説明

DABs Enhancement
Question

What is the supported way for deployment tooling to consume natively resolved, non-secret bundle variables when a bundle declares Unity Catalog secrets?

We also want to understand the compatibility contract for redaction across variable types. Should downstream tools assume that redaction may expand to complex values in a future release?

Observed behavior

Reproduced with Databricks CLI 1.17.0 on Windows, using disposable resources and a dummy secret:

  • Without a UC secret, bundle validate -o json exposes resolved string variable values.
  • With resources.secrets containing a sensitive value, every string variable value becomes [redacted], including unrelated catalog, schema, and warehouse settings.
  • Integer, boolean, and complex values are not handled the same way by the current redaction implementation.
  • bundle deploy still resolves and uses the real values internally.

The string-only check is visible in redact_config.go at v1.17.0.

We understand the reason for protecting variable output. We are asking for a supported integration path rather than a guarantee that type-dependent redaction behavior will remain unchanged.

Minimal example
bundle:
  name: uc-secret-validation-output

variables:
  target_catalog:
    default: example_catalog
  target_schema:
    default: example_schema
  api_credential: {}

resources:
  secrets:
    example_secret:
      name: example_secret
      catalog_name: ${var.target_catalog}
      schema_name: ${var.target_schema}
      value: ${var.api_credential}

targets:
  dev:
    default: true

With an existing workspace profile, using only a dummy value:

$env:BUNDLE_VAR_api_credential = "dummy-value-for-reproduction"
databricks bundle validate --target dev --profile <profile> -o json

The resolved variables.target_catalog.value and variables.target_schema.value are both [redacted]. Removing the UC secret resource makes their normal values visible again. Deployment is not needed to demonstrate the output difference.

Why this matters

Our deployment wrapper uses the same target configuration for pre/post-deployment SQL. It needs the resolved catalog, schema, and warehouse ID, and uses bundle variables when preparing SQL and configuration files.

Reading YAML defaults alone is insufficient: includes, target overrides, environment and file overrides, lookups, and substitutions must agree with native deployment. Reimplementing that resolution logic outside the CLI creates a maintenance and compatibility burden.

Questions for maintainers
  1. Should external consumers treat all type-dependent visibility in validation JSON as an implementation detail that can change with redaction hardening?
  2. Is there a supported command, API, or lifecycle hook for obtaining explicitly selected non-secret resolved values without parsing display-oriented output or changing the bundle configuration?
  3. If not, would you consider such an interface, with an explicit contract for sensitive values and diagnostics?

An existing supported workflow or documentation would also be helpful. We do not need a blanket switch that prints all secrets.

主要言語
Go
スター
409
フォーク
253
平均マージ
1日 11時間
マージ済み PR(30日)
212

環境構築

  • Dockerfile または Docker Compose ファイルあり
  • プルリクエストのテンプレートあり
  • コントリビューションガイドなし

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。

databricks/cli のほかの issue

databricks/cli の issue をすべて見る

似ている issue

Go 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。