Debugging Azure functions locally

Actually the title of this post should be "Debugging your serverless apps locally".
I found a utility while working on Azure functions which helps you debug the function locally(if you have the source code, obviously). It is called ngrok. They have a pretty awesome website and very simple docs.


Steps to proceed are as below.

1. Make sure your Azure function is up and running in your machine and you are able to hit the breakpoint from your local.
2. Signup on the ngrok website.
3. Download the ngrok utility(its an exe file). double click to execute. It opens a command prompt.
4. Connect to your account. Upon login to ngrok website you will get an auth token which you have to use to connect to your account. command to execute "ngrok authtoken <your_token_here>". This step will create a .yml file in your users directory.
5. Create a tunnel. Use command "ngrok http 7071". 7071 is the port you want to expose. I used it because its the default port for Azure function apps.
6. Use the url provided by the ngrok utility to hit the breakpoint on your function from the internet.

I hope this was easy. Please share your thoughts in the comments below.

Getting started with Azure Event Grid

As a human being, every new thing scares us. Same was the case with me when I tried started working on Azure Event Grid. No article I was able to find that works out steps by step. The learning curve looked steep almost vertical ;)
Well in this post I try to reduce those efforts for you. I will try to list out the details required to get you started with Event Grid.

Below image is taken from Microsoft Event Grid documentation website.

Azure Event Grid as its mentioned everywhere, its a way to publish/subscribe your events in an easy way. If your subscriber is down It tries in a Exponential back-off way to deliver the message for 24 hours.

First thing you should know is 'Topic'. What is a topic?
Topic is an endpoint where the source sends the event. To respond to certain types of events, subscribers decide which topics to subscribe to.


Event Grid provides below-mentioned event sources: (Topics are already available, you can subscribe to them easily)

• Azure Subscriptions
• Container Registry
• Custom Topics
• Event Hubs
• IoT Hub
• Media Services
• Resource Groups (management operations)
• Service Bus
• Storage Blob
• Azure Maps

Below mentioned Event Handlers are supported: (You can receive the events using any of the below mechanism)

• Azure Automation
• Azure Functions
• Event Hubs
• Hybrid Connections
• Logic Apps
• Microsoft Flow
• Queue Storage
• Service Bus (Preview)
• WebHooks

The good thing is that you can publish an event of your own using a schema (Custom topic). what it means in real world terms is you call an api with the data in specified format and it will work as event generated and event grid will push it to all the subscribers of that event.

There are three types of schema available.

• Cloud Event Schema
• Custom Event Schema
• Event Grid Schema

In case of inbuilt event providers you can even apply filters to the event, so you will get the event you wish to.

Retry policies can be configured. You can also expire the event after certain time.

Please share your thoughts in the comments below.

Azure functions, A server less compute

First thing I thought about was "What does it mean when someone says its serverless?"

Well I have found my answer serverless doesn't mean it will not run on server first of all. 😉 It means that you don't have to worry about the server resources and other stuff. You just pay for the time your function is executing. If your function executes for 20 seconds per day well then you mostly will pay for that. You  can explore more about the pricing using the links given at the end of this post.

How to get started? Usually that is the most difficult thing to do.

Create a function is as simple as working on a console app. Open Visual Studio. Create new project of Azure functions type and you are done. :)

Make sure you select:

1. Http Trigger
2. Authorization: Anonymous
3. Function Runtime (2.x)
4. Storage Account (Storage Emulator) standalone installer

You are ready to go. Press F5 to run your solution, you can see a command window firing up (which has lot of node.js work behind the scenes). Wait till it gives you the URL which mostly is http://localhost:7071/api/function1if you have not done any changes in the default out of the box app.


There is a pretty comprehensive documentation available.

Azure Functions pricing

 Please share your thoughts in the comments below.

New to KQL? No problem its just like SQL

Well KQL is a query language used to query on Kusto, which is covered in the post Getting started with Kusto.

One Cluster can have multiple Databases and one Database can have multiple Tables. Before going forward let us assume we have a cluster named 'cluster1' and it has database with name 'database1' and database has tables with names 'table1', 'table2', 'table3','table4', 'table5'.

A kusto query usually comprises of  multiple operators and table or cluster name. Unlike SQL table name is not the first thing to be written in KQL. Below is a sample kusto equivalent of select top 5* from table1.

table1 | top 5

Since we are going to query on a big data source lets just add count or where clause to all our queries.

List of  5 basic operators to get started:

1. Project: To select the list of columns to show from the table.

example: table2 | project col1, col2 | take 10.

2. Count: To count the number of rows.

example: table1 | count.

3. Where: It is similar to where clause in SQL server. In KQL we use double equal, unlike SQL, for conditional check.

example: table2 | where col1 = = 'anyvalue'.

4. Take: It picks the number or rows mentioned. It takes rows in no specific order. It is suggested to sort the rows before using take. Limit can also be used for the same purpose.

example: table1 | take 5.

5. Sort (asc/desc): It is used for sorting the data in ascending or descending order, just like SQL.

example: table3 | take 5 col1 asc or table3 | top 5 col1 desc.

You can check out all the available operators/functions/entities at the Microsoft docs website.

Please share your thoughts in the comments below.

Getting started with Kusto

What is kusto?

Well that looks like a difficult question to answer and its not answered clearly (atleast I was not able to find any answer for it). There are many versions of it.Some says its a tool, some says its a query language, docs says its a service. Lets, go with the docs version. Its a service to query over Big Data.

"It is based on relational database management systems, supporting entities such as databases, tables, and columns, as well as providing complex analytics query operators (such as calculated columns, searching and filtering or rows, group by-aggregates, joins)."

What is KQL?
 
KQL is acronym for Kusto Query Language, which can be used to query Azure Application Insight data. There is an open data source available online to query, it can be located at LogAnlytics.



How to Query?

Well to query you first need to learn the basics of KQL but dont worry if you already have worked with SQL or know about T/SQL, its easy to use. The syntax is little similar but there are few new keywords which might need to learn. New way of querying, it feels like mixture of SQL and Powershell to me. All the keywords are separated with Pipes '|' and output from the first query is input to the next one.

"A Kusto query is a read-only request to process Kusto data and return the results of this processing, without modifying the Kusto data or metadata. Kusto queries can use the SQL language, or the Kusto query language."

Where to write Query?

The next part is where can you write the query? Well kusto provides many ways/channels to query the data. All of them are mentioned below. Please select the one that suits your requirement. 😊
As I have mentioned in the article before about LogAnalytics. You can use a web UI to query or you can download Query Explorer. Well you can query over the application insight of your azure subscription also.

There are client libraries available in few languages, There is also an option to connect via Rest API, Client SDK's and KustoExplorer is an available native tool.

    .NET SDK
    Python SDK
    Java SDK
    Node SDK
    PowerShell
    R 

The content for Kusto can be found under names of Azure Data Explorer, Azure Log Analytics etc.
Please share your thoughts in the comments below.