Hello, Siberians. Welcome to lessen 7.4 Off Model seven Off Discuss titled Is that Reserve? One? Microsoft Azure. Hectic Design
Here are the Lenin objectives for this video.
We'll start out by introducing as your event hop. What exactly is the service? What does it do?
Well, then proceeds to cover as your event up from an internal perspective. How does this service to White does out of the function?
We'll cover some con features off Jules Event Hub.
Then we'll do some computers in off event. Message in Service is in Hajer.
Let's get into this.
Let's talk about Joey Vento.
So call it Microsoft Azure Event. Bob is a big data stream in platform on event ingestion. Service
on it can receive and process millions off events for seconds. Data sense to an event up can be transformed and start by using any real time analytics provider or botching storage adapters.
What does that mean? That's the question that I'd like to ask. What does that mean?
Let's talk about joy event hub.
So as your event obviousness, quite bright Microsoft as a big data streaming platform on an event ingestion service. What exactly does that mean,
here's what that means. So in many cases, when we walk with data, the data that we walk, which does not come from a single source
on most likely there many sources that produced data that would need to be ableto unlock in sight. So maybe data coming from high or two devices from sensors from Web and mobile applications that's coming from application telemetry, metrics and then many other sources
on this information usually needs to be available for processing by other systems.
And usually the other systems that also needs to be able to receive and process are analyzed. These data. It's not just one there, multiple off them. So what has your event allows us to do is to be able to ingest data from a variety of sauces on reliably distributes that data
between multiple independent systems
for processing for started on for Ma deep analysis,
let's bring Disarm a little bit
for working with Azure events Harbor The first thing that we need to create It's something called an event ob name space.
After we create as your event up name space, then we can create our event ob within the name space
now where we're creating an event. Hope we need to define something called petitions.
Now the number of petitions that were specified very important because it's not something that we can modify have towards once you've selected a number of petitions, that's number of petition for that event. Hop.
What does the petition corresponds to the number of petitions in an event hop directly relates to the number off concurrent readers that you expects to have.
So that is what that refers to the number of concurrent readers that were intense tohave So, for example, for have multiple event producers that are sending message severe http or am q. P into a jury event hop service
and I have all these different consumer groups, which are the event receivers. They'll be streaming this event into their platforms.
Then I need to ensure that I have the right number off petitions for them to be able to consume those events streaming concurrently.
Some features off as event up number one, as your event of its eyeless calibers can scale up to millions off events per second. That's not the minutes that would in a second
it supports the party CAF occur without the infrastructure management of ahead so as your event, or provide a Kafka and points that can be used by existent, Kafka based applications. So that'll be on as an alternative to you. Run in your own Kafka cluster and building out the infrastructure in managing that
essentially as joy vent. All can be yours
platform As a service managed CAF cluster, you don't have to do with the infrastructure for that kind of jive went off. Support Apache Kafka Protocol 1.0 and later. So if you have existing cuff complications that supports this level of photo Co on and later, you can essentially use actual event hope
to replace your Kafka cluster.
It's a part of you time, much real time. I'm batch processing,
so it's real time. Just case is very interesting because that can never you to stream that directly into a savage that does analytics like I just dream analitico on. Then you can unlock insight into your time.
It supports did a capture to azure blob and agile little aches. As data has been streamed through the service, you can decide to do something called it that a capture on you can capture that street into on a job blob storage, Azure Little League. And then you can go back to review what's been streamed True, the service
it as integrations with dynamics. Tree 65 it as indication with just Dream analytics so we can unlock real time analytics.
Now if you're thinking to yourself,
lot off this service is a You've described your side very, very similar. But the seem tohave published subscribe model where you have some creators that on the left on site and some consumers on divide and site for hold is different service. They look very similar to me. What's the difference between them
that what's the next? The slightest about
so to understand the differences between them with First of all, I need to understand the differences between event on the message. So when we talk about event driven architecture versus message driven architecture, we'll talk about an event, an event, the light wit, notification of a condition of a state change.
What that means is that an event is not necessary leave or data. A good example of that would be if data or foul on objects is going to be dropped into a storage account of blob storage on dhe. I'm gonna trigger on events. I'm going to use an event, a tree, get another execution of another service after that.
Now that event data has information about what happened, but it doesn't have the foul itself. The father was dropped into that blocks of each account is know what the event cavity? Early contents information to say. If our what's dropped, what exactly is inside this foul? It does not have that information.
So, for example, an event notifies consumers that if foul was created,
it may have general information about the foul. Boy doesn't have the found itself. Now compare that to a message.
A message is raw data that produced by a savage to be consumed or soy as well. So it's a message can be the foul itself, like the road to tie itself that's not been produced by a service, not information about an event that are called about the actual body tie itself on. Typically, when you talk about
there's a relationship between the message produce and the message consumer. So where the message producing in some essence as an expectation off what the message consumer would do with that when we talk about events that there's typically no relationship between that. So, for example, if I stopped into a blob storage, I'm going to use that fire off. And as your function
blob storage doesn't have a relationship to say,
Hey, I expect you to do this. It just on events just happens. And then that's pushed out to another service that interested in that event.
So that being said, you can see the differences between the service is at the top there. For event greed, it's supported active programming. Event up is mainly for Big Data Pipeline and for time surveys. Why Service bosses for I Value Enterprise messaging. We're talking about the types off data that the handle
it went great and those events distribution, Which a discreet event.
Why Event Hope and those streaming off a time service event by sufferings like log files vessels, Event event agreed. That's just discreet event on seven's bosses for messages. Raw data on when to use them. Use the rent grade when you want to re hacks to status changes.
He is the vent off when you want to do
telemetry, telemetry and distributed data streaming, especially when you want to do real time streaming for for view time. Another takes on your service, but it brings me to the hand off this lesson. Thanks, right, but for watching and I'll see you in the next lesson.