Azure offers a wide range of services, with which we can build powerful solutions. But how do we know which services to choose, and how to combine them to create even better architectures? In this session, we will take a look at real-life scenarios and how we solved by leveraging the power of Azure.
44. @egrootenboer
Decide on the boundaries of your solutions
Keep pricing in mind when choosing components
Focus on value
Determine your scenario and key constraints
Cloud Solution Architect
Microsoft Azure MVP
InfoQ: Software Development News, Videos & Books
International Speaker
Blogger
Global Integration Bootcamp
Azure IoT Community
Published Author
InfoQ
Boat enthusiast
2
What to expect:
Architecture
Client cases
No Demos
Scalability
Pricing pay as you go
Just bring application
From paas is where the value is
Containers is IaaS++
5
Rustig / duidelijk
Should be based on business and technical requirements
TOFAG, IASA, SAFE
Our experiences, examples of guidelines we see most often
Scalable
Automatic scaling
Technical and organizational
Azure DevOps
GitFlow
Automated testing
ALM
Infrastructure as code
Follow updates
Minimal rights principle
Proactive monitoring
Azure Kubernetes Service
Container based
Full language stack available
Full control
Service Fabric
Large scale
Microsoft stack
First class citizen
Manageability A serverless application is easy to manage, because the platform manages all the of compute resources for you. While an orchestrator abstracts some aspects of managing and configuring a cluster, it does not completely hide the underlying VMs. With an orchestrator, you will need to think about issues such as load balancing, CPU and memory usage, and networking.
Flexibility and control. An orchestrator gives you a great deal of control over configuring and managing your services and the cluster. The tradeoff is additional complexity. With a serverless architecture, you give up some degree of control because these details are abstracted.
Portability. All of the orchestrators listed here (Kubernetes, DC/OS, Docker Swarm, and Service Fabric) can run on-premises or in multiple public clouds.
Application integration. It can be challenging to build a complex application using a serverless architecture. One option in Azure is to use Azure Logic Apps to coordinate a set of Azure Functions. For an example of this approach, see Create a function that integrates with Azure Logic Apps.
Cost. With an orchestrator, you pay for the VMs that are running in the cluster. With a serverless application, you pay only for the actual compute resources consumed. In both cases, you need to factor in the cost of any additional services, such as storage, databases, and messaging services.
Scalability. Azure Functions scales automatically to meet demand, based on the number of incoming events. With an orchestrator, you can scale out by increasing the number of service instances running in the cluster. You can also scale by adding additional VMs to the cluster.
Manageability A serverless application is easy to manage, because the platform manages all the of compute resources for you. While an orchestrator abstracts some aspects of managing and configuring a cluster, it does not completely hide the underlying VMs. With an orchestrator, you will need to think about issues such as load balancing, CPU and memory usage, and networking.
Flexibility and control. An orchestrator gives you a great deal of control over configuring and managing your services and the cluster. The tradeoff is additional complexity. With a serverless architecture, you give up some degree of control because these details are abstracted.
Portability. All of the orchestrators listed here (Kubernetes, DC/OS, Docker Swarm, and Service Fabric) can run on-premises or in multiple public clouds.
Application integration. It can be challenging to build a complex application using a serverless architecture. One option in Azure is to use Azure Logic Apps to coordinate a set of Azure Functions. For an example of this approach, see Create a function that integrates with Azure Logic Apps.
Cost. With an orchestrator, you pay for the VMs that are running in the cluster. With a serverless application, you pay only for the actual compute resources consumed. In both cases, you need to factor in the cost of any additional services, such as storage, databases, and messaging services.
Scalability. Azure Functions scales automatically to meet demand, based on the number of incoming events. With an orchestrator, you can scale out by increasing the number of service instances running in the cluster. You can also scale by adding additional VMs to the cluster.
30
Rustig / duidelijk
Logic Apps
Over 250 connectors
Visual designer
Rich monitoring and management
Blob Storage
Video Indexer
Azure Media Services
Media Encoding
Durable Functions
Execute code in several languages
Portable runtime
Serverless or App Service Plan
Azure Automation
Operations focus
Primarily PowerShell
Windows and Linux
Build / Deploy resources - Deploy VMs across a hybrid environment using Runbooks and Azure Resource Manager templates. Integrate into development tools like Jenkins and Azure DevOps.
Configure VMs - Assess and configure Windows and Linux machines with the desired configuration for the infrastructure and application.
Monitor - Identify changes on machines that are causing issues and remediate or escalate to management systems.
Protect - Quarantine VM if security alert is raised. Set in-guest requirements.
Govern - Set up role-based access control for teams. Recover unused resources.
Build / Deploy resources - Deploy VMs across a hybrid environment using Runbooks and Azure Resource Manager templates. Integrate into development tools like Jenkins and Azure DevOps.
Configure VMs - Assess and configure Windows and Linux machines with the desired configuration for the infrastructure and application.
Monitor - Identify changes on machines that are causing issues and remediate or escalate to management systems.
Protect - Quarantine VM if security alert is raised. Set in-guest requirements.
Govern - Set up role-based access control for teams. Recover unused resources.
40
Rustig / duidelijk
Skip 53
So, with that, let’s have a quick talk about devops.
Not just technical, also business