You need to enable JavaScript to run this app.
优惠活动
大模型
产品
解决方案
定价
更多

Azure SQL DB高级版只读缩放:能否指定连接特定只读副本?

回答

Great question! Let me break down how this works with Azure SQL Database's read scale-out feature for you:

  • First, you’re right that Premium/Business Critical tier databases with Read Scale-out enabled come with read-only replicas, and adding ApplicationIntent=ReadOnly; to your connection string routes traffic to these replicas. However, there’s no built-in way to directly specify connecting to "Replica1" or "Replica2" specifically.

  • The read-only replicas in Read Scale-out are managed as a shared pool, and Azure uses an internal load balancer to distribute incoming read requests across them automatically. This means you can’t target a single replica via connection string parameters—your requests will be balanced across all available replicas in the pool.

  • If you need to isolate different workloads (like running analytics on one dedicated replica and Power BI reports on another), a better alternative is to use geo-replicated read-only replicas. Each geo-replicated replica has its own unique endpoint, so you can create separate connection strings for each workload:

    • Use the first geo-replica’s endpoint for your analytics tasks
    • Use the second geo-replica’s endpoint for Power BI reports

Here’s a quick reminder of the standard connection string for load-balanced access to the Read Scale-out replica pool:

Server=tcp:<your-server-name>.database.windows.net;Database=<your-db-name>;ApplicationIntent=ReadOnly;User ID=<your-username>;Password=<your-password>;Trusted_Connection=False; Encrypt=True;

To sum up: Read Scale-out’s built-in replicas are shared and load-balanced, so targeting individual ones isn’t possible. For workload-specific dedicated replicas, geo-replicated instances are the recommended solution.

内容的提问来源于stack exchange,提问作者Oleh

相关产品推荐
方舟 Agent Plan

超全模态模型 × Harness 升级,最新支持 Deepseek-V4.1-Flash、GLM-5.3 系列、Doubao-Seedream-5.0-pro、Kimi-K3 (部分), 限时 9.9 元起

最近更新时间:2026.05.13 07:42:52