Sample - Customer Reviews App with Cognitive Services (Azure Functions tools for Visual Studio 2017)

Dernière mise à jour : 27/06/2017
Modifier sur GitHub

This sample showcases the new Azure Functions tooling for C# Class Libraries in Visual Studio. It has a website where customers submit product reviews, stored in Azure storage blobs and CosmosDB; and an Azure Function to perform automated moderation of the customer reviews using Microsoft Cognitive Services. It uses an Azure storage queue to decouple the website from the function.

Setup

Please follow the Setup Instructions to get your environment configured.

Sample Highlights

Sample Architecture

Here's a view of the architecture of the sample:

Deliver as a Demo

Here is the demo script.

And here's an example recording of the demo:

Section 1: Azure Functions tools for Visual Studio 2017

This section highlights the benefits of creating Azure Functions in Visual Studio 2017: compiled class libraries, attribute based function definition, intellisense, live unit testing, local debugging.

| Screen | Steps | Notes | |-------------- |Right click on the in Visual Studio and select Commit.

Select the commit textbox and enter a short commit text (or click Ctrl+Shift+c to paste the commit code if using AutoHotKey).

Click on Commit and Sync|Without even leaving your development environment you'll commit these changes to your repo and trigger a build.| | |After a while, open Team Explorer - Builds in Visual Studio Team Services.

Right click on latest build and select Open|Commiting the code will trigger an existing continuous integration pipeline that you created with VSTS to build and test your changes. In a real scenario you would also submit a pull request from Visual Studio or in VSTS and let your function code get reviewed by your team, and once merged have it be deployed by VSTS to test and other environments using Release Management.

If you want to learn more about this build run you can right click and open it from Visual Studio| | |Check out the build running/or finalized if done already in VSTS|Here you can see your build is successful, passed the unit tests, and your latest code was pushed up to your dev Azure Function by VSTS. You have this solution already deployed to Azure, and now you can close the loop and see how users are using it and how it is behaving.|

Section 3: Application Insights

This section highlights how Application Insights custom telemetry can be surfaced on Azure along with overall telemetry from the demo website.

| Screen | Steps | Script | |-------------| ||Check out the code the EmitCustomTelemetry method to see how the custom telemetry is being sent.

Open your Azure Portal tab with the Customer Reviews cusom dashboard in the browser|In addition to the best development tooling and great integration with Visual Studio Team Services, you can take advantage of Application Insights to easily build custom telemetry into your application. You will see that with a few lines of code you're collecting and storing application telemetry for deep insights and analysis.| | |Mouse over the donut wheel in the shared Reviews Insights dashboard.

Hover over a couple of areas of the donut to see their values.

Hover over the application map and interact with it.

Click on the edit query button in the donut to get to the Application Insights Analytics portal|When you go to the Azure portal you have rich visualizations and deep insights about how your app and your function are performing in a shared dashboard. For example you can see how many Reviews have been approved or rejected and why. You can also see how many calls are being made to your site, Azure Function, and their dependencies, and the latency of these calls. And if you need to you can drill into the specifics of the telemetry using Application Insights Analytics| | |You'll see type of query result changed to barchart

You can click Go to run the query again and see the results as barchart.|Here you can drill in to individual requests, aggregate your results in a different way. You can change the view type and re-run the query. You can export the new view to your Azure Portal dashboard, or use this query as a source in Power BI.

So, with Azure Functions, Visual Studio, Team Services, and App Insights you and your team have the best tooling for creating, deploying and monitoring serverless solutions.|

Contributing

This project has adopted the Microsoft Open Source Code of Conduct. For more information see the Code of Conduct FAQ or contact opencode@microsoft.com with any additional questions or comments.

Autres exemples par Thiago Almeida