最佳做法要求:一个或多个appinsight帐户? [英] Best practices request: One or many appinsight accounts?

问题描述

亲爱的Appinsights专家

Dear Appinsights Pros,

我们正在开发许多Web应用程序，conatiner，API等.而且，我们希望借助appinsights获得有关解决方案的见解

We are developing a lot of web apps, conatiners, APIs and so on. And we like to use appinsights to gain insights on the solutions  

最近在团队中进行的讨论是:我们应该将所有API再次链接到一个应用程序洞察力帐户还是应该为每个API添加一个appinsight?

A discussion in the team that arose these days was: should we link all APIs agains a single app insight account or should we add an appinsight per API?

如果推荐第二个，我们如何获得所有Appinsight的概述，以便在一种伞状仪表板中查看它们?

If the second is recommended  how to we get an overview over all appinsights to see them in a kind of umbrella dashboard  

感谢分配

亲切的问候

Benedikt

Benedikt 

推荐答案

你好Benedikt

Hello Benedikt,

选择一个或多个资源将取决于您尝试为其设置遥测的系统的实现.

Choosing one vs. many resources will depend on the implementation of the system you're trying to setup telemetry for.

例如，如果其中一些API彼此通信/通过HTTP相互代理，那么AI会自动为您关联这些API.在这种情况下，通过打开任何AI资源并检测到问题/运行速度减慢，您将看到端到端事务 跨过所有涉及的API，无论其数据位于哪个AI资源中(当然，如果您有权访问Azure Portal中的那些资源).这是这种分布式的例子

For instance, if some of these APIs talk to one another / proxy one another via HTTPs, then AI will automatically correlate those for you. By opening any AI resource with an issue / slowdown detected in such a case, you'll see end-to-end transaction that crossed all involved APIs regardless of which AI resource their data landed in (if you have access to those resources in Azure Portal, of course). Here is the example of such distributed transaction and map views.

或者，如果这些资源中的大多数是独立的并且不会互相交谈，那么要拥有一个单一的玻璃窗格，您可以选择:

Alternatively, if most of these resources are independent and won't talk to one another, then to have a single pane of glass you may either go with:

-用于所有API的单一资源，然后通过对API名称进行切片来通过Metrics Explorer或Analytics(分析)查询来构建Azure仪表板(或AI工作簿)；

- Single resource for all APIs and then build Azure Dashboards (or AI Workbooks) via Metrics Explorer or Analytics queries by slicing on API name;

-多种资源(每个API一个)，然后通过切片和重新命名来通过ME和Analytics(分析)查询构建Azure仪表板(或AI工作簿).或者，只需使每个图表代表其自己的资源即可.

- Multiple resource (one per API) and then build Azure Dashboards (or AI Workbooks) via ME and Analytics queries by slicing on resoruce name instead. Or, simply make each chart represent its own resource.

这篇关于最佳做法要求:一个或多个appinsight帐户?的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！