And here we are. The first proper post.
I’ve found over the years that the best way to get a grip of how azure resources work is by building an end to end solution. You get to understand how data is consumed, processed and presented. With that also comes different security configurations, limitations, understanding of costs and how extendable the resources can become.
Tweet analysis is always popular, and the Microsoft Example is pretty thorough but I had never seen anything where all the pieces are strung together and it that case a lot of the legwork is done and connectivity between some of the resources is hidden to you, so I thought what better way to kick this blog off than to go through that solution. I’ll break it in down into the following parts to make it digestible
Part 1: The Overview
Part 2: The Twitter Account, App & and Event Hub
Part 3: Streaming Analytics and Power BI
We need to get the tweets from somewhere, right? We’ll be opening up a stream of tweets that will search for a given keyword. In this case we’ll look for tweets about Microsoft. We’ll need to register an app and retrieve some keys, but more of that in Part 2.
The event hub will ingest and stream our tweets to supported resources within azure. The way it differs from other messaging services that it’s has the ability to consume millions of events per second, maintains orders of tweets and can also capture the events in an associated blob (configurable at setup).
Allows us to plug into the stream of tweets and perform some analysis through querying or calling machine learning workbench or windowing function. It can seamlessly connect with numerous resources and in our case, Event hub and Power BI.
We’ll be presenting a live dashboard of the tweet analysis in Power BI. It’s easy to get up and running with it and create what seems like complex visuals in a matter of minutes.
That’s the little intro.
Now let’s get to Part 2 and get our hands dirty.