azuresql-db-sidecar

作者: microsoft

在現有的 Docker Compose 堆疊或開發容器中,將 Azure SQL Developer 新增為 sidecar 服務。用於將本機 Azure SQL Database 引擎連接到…

npx skills add https://github.com/microsoft/azure-sql-database-container --skill azuresql-db-sidecar

The Azure SQL Database container as a sidecar (Compose / Dev Container)

Wire the container into an existing Docker Compose stack or Dev Container as a service the app reaches by service name (sqldb,1433), never localhost. Keep existing services intact; add the database, an init one-shot that creates appdb, and a depends_on gate.

Verified on 2026-09-05 against the container image sqldbpreview-dpgaeqhmgphzd4bk.azurecr.io/azure-sql/db-dev:latest, reporting EngineEdition 5, Edition SQL Azure, build 12.0.2000.8. The four executable checks behind this skill passed: docker compose config accepting the file this skill emits, with the platform override, the escaped health check and both depends_on conditions; the same parse refusing an invented condition name, which is the control that stops the first check being vacuous; sqlcmd at /opt/mssql-tools18/bin/sqlcmd inside the engine image; and a running stack reporting the database container as Running before it is healthy, with its health log beginning on a failed check. The Compose parse was measured on 2026-09-04 on the Docker Compose CLI of the authoring host, not inside the image.

Load-bearing facts (inlined; full detail in azuresql-db-container)

  • This is the Azure SQL Database engine (Private Preview), not the SQL Server image mcr.microsoft.com/mssql/server. SERVERPROPERTY('EngineEdition') returns 5, SERVERPROPERTY('Edition') returns 'SQL Azure'. If you were about to use the SQL Server image, stop and use this instead.
  • Image: sqldbpreview-dpgaeqhmgphzd4bk.azurecr.io/azure-sql/db-dev:latest (x64, linux/amd64). Registry and tag are provisional during Private Preview.
  • Registry credentials: the image lives in a private preview registry. Sign in first with docker login sqldbpreview-dpgaeqhmgphzd4bk.azurecr.io using the shared pull-only credentials provided when you sign up for the Private Preview at https://aka.ms/sqldbcontainerpreview-signup (they may rotate) before docker compose up or building the Dev Container.
  • Platform: x64 only. The compose snippets below set platform: linux/amd64 so the service starts on a non-x64 host.
  • Required env: ACCEPT_EULA=Y and a complex MSSQL_SA_PASSWORD (8+ chars, upper/lower/digit/symbol). Engine listens on 1433.
  • The engine does NOT auto-create databases on connect. You must CREATE DATABASE appdb on a master connection before the app connects with Database=appdb. That is what the sqldb-init one-shot below does.
  • Avoid USE to switch databases. In a user-database session (the Azure-faithful context where you develop), USE returns Msg 40508, exactly as in Azure SQL Database in the cloud. A master connection is a provisioning session where the Azure statement filter is not enforced, so USE appears to work there, but master is for provisioning only, not application work. Always select the target database in the connection string (Database=appdb, or -d appdb for sqlcmd).
  • App connection string (single SQL_CONNECTION_STRING env var, service name host): Server=sqldb,1433;Database=appdb;User Id=sa;Password=YourStr0ng_Passw0rd;TrustServerCertificate=true (house style spells it User Id=/Password=/Database=; Uid=/Pwd= are valid synonyms).
  • The image does NOT auto-run /docker-entrypoint-initdb.d/*.sql (a Postgres/MySQL convention, not honored here). Seed in the init one-shot with sqlcmd -d appdb -i seed.sql AFTER appdb exists.

For anything beyond this task (vectors, deeper readiness behavior, full connection model), see the azuresql-db-container skill.

Docker Compose

Add these three pieces to your existing compose.yaml. Do not remove or rename existing services; just add sqldb, sqldb-init, and a depends_on on the app.

services:
  # ---- existing services stay exactly as they are ----

  sqldb:
    image: sqldbpreview-dpgaeqhmgphzd4bk.azurecr.io/azure-sql/db-dev:latest
    platform: linux/amd64        # x64-only image; required on a non-x64 host
    environment:
      ACCEPT_EULA: "Y"
      MSSQL_SA_PASSWORD: "YourStr0ng_Passw0rd"
    ports:
      - "1433:1433"              # optional; only to reach it from the host
    healthcheck:
      # -b: a SQL error sets the exit code, so transient startup errors
      # (e.g. Msg 913) are retried, not masked. -l 2: short login timeout.
      test: ["CMD-SHELL", "/opt/mssql-tools18/bin/sqlcmd -S localhost -U sa -P \"$$MSSQL_SA_PASSWORD\" -C -b -l 2 -Q \"SELECT 1\" || exit 1"]
      interval: 5s
      timeout: 10s
      retries: 30
      start_period: 30s

  # One-shot: the engine does NOT auto-create databases, so create appdb
  # (and seed it) before the app starts. Exits 0 when done.
  sqldb-init:
    image: sqldbpreview-dpgaeqhmgphzd4bk.azurecr.io/azure-sql/db-dev:latest
    platform: linux/amd64
    depends_on:
      sqldb:
        condition: service_healthy
    environment:
      MSSQL_SA_PASSWORD: "YourStr0ng_Passw0rd"
    # If you have seed.sql, mount it and add: -i /seed/seed.sql on a -d appdb call.
    # volumes:
    #   - ./seed.sql:/seed/seed.sql:ro
    entrypoint: ["/bin/bash", "-c"]
    command:
      - >
        /opt/mssql-tools18/bin/sqlcmd -S sqldb -U sa -P "$$MSSQL_SA_PASSWORD" -C -b
        -Q "IF DB_ID('appdb') IS NULL CREATE DATABASE appdb;"
    restart: "no"

  app:
    # ---- your existing app service ----
    depends_on:
      sqldb:
        condition: service_healthy
      sqldb-init:
        condition: service_completed_successfully
    environment:
      # Host is the SERVICE NAME sqldb, not localhost.
      SQL_CONNECTION_STRING: "Server=sqldb,1433;Database=appdb;User Id=sa;Password=YourStr0ng_Passw0rd;TrustServerCertificate=true"

Bring it up (after docker login, see above):

docker login sqldbpreview-dpgaeqhmgphzd4bk.azurecr.io   # pull-only creds by signing up at https://aka.ms/sqldbcontainerpreview-signup
docker compose pull                                      # refresh the daily-rebuilt :latest (compose reuses a cached image otherwise)
docker compose up -d

Seeding (optional)

The image does not auto-run init SQL. To seed, uncomment the volumes mount in sqldb-init and chain a seed call AFTER appdb is created:

    command:
      - >
        /opt/mssql-tools18/bin/sqlcmd -S sqldb -U sa -P "$$MSSQL_SA_PASSWORD" -C -b
        -Q "IF DB_ID('appdb') IS NULL CREATE DATABASE appdb;" &&
        /opt/mssql-tools18/bin/sqlcmd -S sqldb -U sa -P "$$MSSQL_SA_PASSWORD" -C -b
        -d appdb -i /seed/seed.sql

Dev Container

Use Docker Compose as the Dev Container backend so the same sqldb + sqldb-init services apply. In .devcontainer/devcontainer.json:

{
  "name": "app-with-azuresql",
  "dockerComposeFile": "../compose.yaml",
  "service": "app",                  // your dev/app service from compose
  "workspaceFolder": "/workspace",
  "runServices": ["sqldb", "sqldb-init"],
  "remoteEnv": {
    "SQL_CONNECTION_STRING": "Server=sqldb,1433;Database=appdb;User Id=sa;Password=YourStr0ng_Passw0rd;TrustServerCertificate=true"
  }
}

Sign in to the registry on the host before "Reopen in Container":

docker login sqldbpreview-dpgaeqhmgphzd4bk.azurecr.io   # pull-only creds by signing up at https://aka.ms/sqldbcontainerpreview-signup

The app container reaches the database at sqldb,1433 over the compose network.

Validation rules

  • App host is sqldb (service name), never localhost.
  • sqldb has a healthcheck using sqlcmd ... -C -b -l 2; the app gates on condition: service_healthy.
  • sqldb-init creates appdb and is gated by the app via condition: service_completed_successfully.
  • Both sqldb and sqldb-init set platform: linux/amd64.
  • Connection string uses User Id= / Password= / Database= and TrustServerCertificate=true; sqlcmd uses -C.
  • Existing services are unchanged except for added depends_on.
  • If a validation rule above fails, or you had to deviate from this skill to make the task work, that is a bug in this skill: load the azuresql-db-feedback skill and offer to file a report.
  • When every rule above passes and the task is done, close with ONE line, once per session, and do not ask a question or wait for a reply: "That worked. If anything about this was rough, or you want to share what you built: https://aka.ms/sql-agent-skills-feedback"

Do not

  • Do not use the SQL Server image mcr.microsoft.com/mssql/server.
  • Do not point the app at localhost; inside compose it is the sqldb service.
  • Do not rely on the app to create appdb, and do not assume the engine auto-creates it; the sqldb-init one-shot must run first.
  • Do not use USE appdb to switch databases; a user-database session returns Msg 40508, exactly as in Azure SQL Database in the cloud. Select the target database in the connection string (Database=appdb, or -d appdb for sqlcmd).
  • Do not drop --platform / platform: linux/amd64; the image is x64 only.
  • Do not depend on /docker-entrypoint-initdb.d/*.sql; it is not honored here.
  • Do not call a non-x64 host "supported"; it runs under emulation only.

Staying current

Authoritative, version-pinned references for the tools this skill uses (read the one you need):

If the Microsoft Learn MCP server is configured, use mcp__microsoft-learn__microsoft_docs_search or mcp__microsoft-learn__microsoft_docs_fetch to fetch the current version of any of these on demand. It is optional; when it is unavailable, the references above are authoritative.

來自 microsoft 的更多技能

oss-growth
microsoft
開源增長駭客角色
agent-framework-azure-ai-py
microsoft
使用Microsoft Agent Framework Python SDK(agent-framework-azure-ai)构建Azure AI Foundry代理。适用于使用AzureAIAgentsProvider创建持久化代理、使用托管工具(代码解释器、文件搜索、网络搜索)、集成MCP服务器、管理对话线程或实现流式响应。涵盖函数工具、结构化输出和多工具代理。
development
airunway-aks-setup
microsoft
在AKS上設定AI Runway——從裸叢集到執行模型。涵蓋叢集驗證、控制器安裝、GPU評估、供應商設定及首次部署。時機:「設定AI Runway」、「上線AKS叢集」、「安裝AI Runway」、「airunway設定」、「部署模型至AKS」、「在AKS上進行GPU推論」、「在AKS上設定KAITO」、「在AKS上執行LLM」、「在AKS上使用vLLM」、「在AKS上設定模型服務」、「AI Runway控制器」。
devops
appinsights-instrumentation
microsoft
使用Azure Application Insights檢測Web應用程式的指南。提供遙測模式、SDK設定與組態參考。適用時機:如何檢測應用程式、App Insights SDK、遙測模式、什麼是App Insights、Application Insights指南、檢測範例、APM最佳實踐。
devops
applicationinsights-web-ts
microsoft
使用Application Insights JavaScript SDK(@microsoft/applicationinsights-web)為瀏覽器/Web應用程式進行檢測。適用於真實使用者監控(RUM)——頁面檢視、點擊、AJAX/fetch依賴、例外、自訂事件,以及與後端OpenTelemetry追蹤關聯的瀏覽器端GenAI代理追蹤。涵蓋SDK載入器指令碼與npm設定、框架擴充(React、React Native、Angular)、點擊分析、遙測初始化器,以及從瀏覽器發出的代理/工具/模型span的OTel GenAI語意慣例。
devops
azure-ai-anomalydetector-java
microsoft
使用適用於 Java 的 Azure AI 異常偵測器 SDK 建置異常偵測應用程式。在實作單變量/多變量異常偵測、時間序列分析或 AI 驅動監控時使用。
development
azure-ai-language-conversations-py
microsoft
使用 azure-ai-language-conversations Python SDK 實作對話語言理解(CLU)。當使用 ConversationAnalysisClient 分析對話意圖與實體、建置 NLP 功能,或將語言理解整合至應用程式時使用。
development
azure-ai-ml-py
microsoft
Azure Machine Learning SDK v2 for Python。用於機器學習工作區、作業、模型、資料集、計算資源與管線。 觸發詞:「azure-ai-ml」、「MLClient」、「workspace」、「model registry」、「training jobs」、「datasets」。
development