Hello Austin, hello Azure Developer Tour!
Past April 11th I had the opportunity to attend the Azure Developer tour in Austin TX. It was a nice event at the ZACH Theater in this charming city. The event was one more of the tour that Microsoft has been doing to take Developer Advocates closer to the communities. The event had around 200 attendees, and was hosted by Maxime Rouiller (@maximrouiller) and Ruth Yakubu (@ruthieyakubu), who did some great presentations and live demos showcasing some of the Azure products in action with real world applications.
I already knew about most of the products, but no doubt there were new features that woke up my curiosity, and I went back to my home willing to try the new things and do more coding using the Azure platform.
Below I summarized the main features and highlights I took about each one of products showcased.
Service Fabric
Infrastructure managed by the platform. Its usage is for development of microservices and container-based applications. In this product, applications are auto-scalable, stateless or stateful, and integrated with Azure SQL Database, Azure Cosmos DB, Cortana, Microsoft Power BI, Microsoft Intune, Azure Event Hubs, Azure IoT Hub, Dynamics 365 and Skype for Business among others. Also able to mix services in processes and services in containers in the same application. Container orchestrator. Integration with VSTS, Jenkins, Octopus Deploy and others CI/CD tools.
Service Fabric can be run on premise, thus no risk of lock-in to the provider but lock-in to the platform.
Link to official documentation here
App Service
Platform to deploy a complete traditional application. Not a target for microservices, although the amount of functionality per application remains up to the developer. No Platform Lock-in. Control over the web.config and the applicationhost.config in app services, which should be used with caution, the more is customized the more it needs to be maintained.
From the developer point of view is pretty easy to get started. Integration within the platform with authentication providers, automatically creates and integrates application insights for monitoring.
Data should be stored in an external datastore like CosmosDB or Azure SQL.
Link to official documentation here
Azure Functions
My favorite one, serverless compute service to run code on-demand without having to explicitly provision or manage infrastructure. Run code in response to a variety of events like: Time, Data Processing and Webhook+API’s. Obviously tide to the platform and very easy to get started.
Stateless only. Link to official documentation here.
Examples of functions:
- Create a function that runs on a schedule
- Create a function triggered by Storage queue messages
- Create a function triggered by a generic webhook
- Create a function triggered by a GitHub webhook
- Add messages to an Azure Storage queue using Functions
- Store unstructured data in Azure Cosmos DB using Functions
Microsoft cognitive services
Mainly Microsoft Azure Machine Learning Studio, about predictive analytics, for data scientists. Cortana Services. The Azure Experiments tool was also shown, which is a selection of components to create predictions based on sets of data. Saved Datasets and management. R Language and Python support. Vision, Language, Speech, Knowledge, and Search API’s. Link to documentation here.
Logic Apps
Azure services, Microsoft services, and other software-as-a-service (SaaS) apps integrated in a customized and automated workflow that triggers scheduled. A more evolved version of Azure Functions. Some of the connectors are Office, OneDrive, Oracle Database, Salesforce, SAP, Sharepoint, Trello, Twitter, SQL Server, Slack among others. Can be developed from VS with the Azure SDK. A quick tutorial can be seen here
CosmosDB
Besides being a document database, it allows to store graphs and Key Value pairs. It can be used along with SQL API, Graph API, Table API, MongoDB API, or Cassandra API. Here is a quick start tutorial to get an idea of what you can do. Documentation here.
PostgreSQL and MySQL in Azure
Both databases are now dataservice in Azure with General Availability. Link to documentation to both products here and here.
Conclusion
Overall the experience of attending the event was great. Developers were the fundamental target of the Microsoft Azure staff at the event. With the promise that the developer experience when getting to know the platform will be easier and, nicer. The overview of each product was well done with live demos of real use cases. e.g. collecting data from twitter, creating graphs, and analytics. Some of the products in the Azure cloud are providing a better development experience, due the integration of Visual Studio with the product. The offer of the products is focused more on the services where the infrastructure is managed by the platform, so the user/developer can focus only in the business value of the solutions implemented in the platform. The biggest announcement was the Machine Learning Studio, which offers tools for AI and is well integrated with the platform. I only missed probably a quick hands on lab at the event, but definitely looking forward to more events and keep learning with all the resources learned there. Thanks Microsoft!
