Hybrid cloud computing is the combination of a public cloud provider (such as Azure) with on-premise infrastructure and assets. For organizations looking to optimize, a hybrid cloud solution offers the best of both on-premise and cloud computing worlds. The Azure platform is rich with services and applications that can help organizations leverage new technology while keeping their assets local. This article takes a look at some of the different dimensions of an Azure approach to hybrid cloud computing.
Hybrid Cloud Infrastructure
In a hybrid cloud infrastructure approach, customers use Azure to deploy servers and platforms within an extended network connected to their local domain. Using the Azure VPN Gateway, organizations can connect to a Virtual Network in the cloud using a site-to-site gateway and treat it as an extension of their own network. Individuals can also be given point-to-site access to the virtual network using the same technology.
in this configuration, you can create Windows or Linux Virtual Machines built to your own specifications, implement a custom commercial platform (with built-in licensing) from the Azure Marketplace, or launch an open source platform VM from the VM Depot, an oft-overlooked Azure catalog.
In traditional on-premise environments, businesses leveraged IIS servers running ASP.NET, WCF and web services for internal and external applications; and BizTalk Server for enterprise integration with internal and external business partners using EDI, XML and other formats across various messaging and data transfer protocols. In this particular model, Azure has arguably provided the most game-changing innovation for hybrid cloud solutions that exists in the market among vendors today.
Microsoft has taken three application models – web, app, and business process – and given them a unified base called the App Service. There are four general types of App Service apps you can create:
- Web Apps for scalable website and web applications
- Mobile Apps for device-neutral application development
- API Apps for creating and consuming APIs
- Logic Apps for automating data access and usage across clouds and platforms
As a single integrated service, App Service makes it easy to compose the above app types into a single solution, allowing you to easily build apps that target both web and mobile clients using the same back-end and integrate with on-premise systems as well as popular services such as Office 365 and Salesforce.
Apps and Connectors
Azure App Services is built atop a principle of extensibility and common connectivity through API Apps. A Connector is a type of API App that focuses on connectivity. Connectors, like any other API App, are used from Web Apps, Mobile Apps, and Logic Apps. Connectors make it easy to connect to existing services and help manage authentication, provide monitoring, analytics, and more. Any developer can create their own API Apps and deploy them privately. In the future, developers can share and monetize their custom-created API Apps through the marketplace.
There are two tiers of connectors available – Standard and Premium. As noted in the List of API Connectors and API Apps to use in your Logic Apps, functionality that used to be provided in BizTalk – EDI X12, EDIFACT, AS2, BizTalk Transform Service, BizTalk Trading Partner Management, Business Rules – are now Premium Connectors for Logic Apps. These connectors can be used as poll/push triggers to initiate action on the receipt of actionable data; or as actions, allowing CRUD operations or doing data lookups as part of another step in a process. Message logging, archiving, and diagnostics logging are all built in.
When BizTalk Server was initially implemented in Azure, it was called Windows Azure BizTalk Services and contained all of the components people would expect to find in a typical BizTalk solution. As you’ve seen, functionality has been broken up into consumable Connectors that can be used as needed in the App Service model, but that was only half of it. The other half of the platform allowed customers to deploy a local agent into their on-premise environment and serve as an agent for feeding data back and forth to the cloud. This piece has now been redeployed as Hybrid Connection.
The Hybrid Connections feature of what used to be BizTalk Services lets you connect Azure Apps to any on-premises TCP or HTTP resource—such as Microsoft SQL Server, MySQL, or any web service—with just a few configuration changes and without using any custom code.
Applications using Hybrid Connections access only the specific on-premises resource that is published through the Hybrid Connection.
For message-based services and applications, the Azure Service Bus provides four different communication mechanisms:
- Queues for one-directional communication to single recipients
- Topics for one-directional communication to multiple recipients, also known as subscribers
- Relays for bi-directional communication. Unlike Queues or Topics, Relays are simply passed on to the destination application.
- Event Hubs which provide event and telemetry data to the cloud at massive scale, with low latency and high reliability.
Hybrid Data Solutions
For organizations interested in performing tasks with their data without the actual relocation of said data into cloud-based storage, a number of rich services have appeared.
Auzre Data Factory gives you an agile way to manage the production of trusted information from complex processes, making it easier to create, orchestrate, and manage data-processing pipelines over a range of transformation services and diverse data sources. Many of the data sources supported by the Data Factory are on-premise, allowing you to move data from one system to another in an ETL pattern.
A full catalog of the supported on-premise data stores, see Data movement activities.
Microsoft Azure Data Catalog is a fully managed service hosted in the Microsoft Azure cloud that serves as a system of registration and system of discovery for enterprise data sources. Azure Data Catalog provides capabilities that enable any user – from analysts to data scientists to developers – to register, discover, understand, and consume diverse and home-made/manual cloud and on-premise data sources.
Currently the data can only be consumed by Excel and/or SQL Server Reporting Services. Some of the common scenarios listed for using this service include:
- Registration of central data sources. These data sources include line of business OLTP systems, data warehouses, and business intelligence / analytics databases. Often the number of systems, and the overlap between systems, grows over time as the needs of the business evolve, and as the business itself evolves through mergers and acquisitions. For customers getting started with Azure Data Catalog, identifying and registering the key data sources used by many different teams of data consumers can be the first step to success.
- Self-service business intelligence. Self-service BI allows information workers and analysts to create their own reports, workbooks, and dashboards without relying on a central IT team – or being restricted by that IT team’s schedule and availability.
- Capturing tribal knowledge. Point to (SharePoint, for example) locations of data repositories used for certain purposes, such as employee onboarding. You can annotate the the data assets and include links to existing documentation.
Azure Data Catalog is currently in Preview at the time of this writing.
Microsoft Power BI is a business intelligence suite that delivers dashboards and reports using datasets, which (wait for it) can be cloud-based or on-premise. You can also pull data from Microsoft’s Platform-as-a-Service (PaaS) offerings, such as Office 365 Excel files in OneDrive or SharePoint data in SharePoint Online.
The Azure services stack has a rich and growing catalog of services designed to help organizations grow their capabilities outside of the traditional on-premise server platform model. Next up, we’ll take a look at Microsoft’s newest approach to hybrid cloud computing in the Azure platform: the Azure Stack.
Images courtesy of Microsoft.