Performance and Schedule Autoscaling in Microsoft Azure
Do you have a website or application that experiences heavy traffic in peaks during certain days, weeks or months of the year? Do your servers sit idle for extended periods of time? Are you paying to support your peak loads in off-peak times? If you answered “Yes” to any of these questions, Azure autoscaling may be something to consider. Read my blog on performance and schedule autoscaling in Azure to learn more.
Azure is Microsoft’s cloud platform for Infrastructure as a Service (IaaS) and Platform as a Service (PaaS). Comprised of a growing collection of integrated services for applications, computing, data, networking and storage, Microsoft Azure has been ranked as an industry leader by Gartner due to its ability to help organizations do more, work faster and save money. Azure’s IaaS offering gives you the flexibility to define your environment requirements based on assumptions, then make adjustments once those assumptions are confirmed – in minutes. For example, if you expect 1,000 concurrent users to hit your site for the entire month of December during the peak shopping season, but later find that you only need to support that high load during the third week of December, Azure can be configured to dynamically adjust the number and size of servers as needed.
There are essentially two types of scaling options in Azure:
- Performance-based Autoscaling
- Schedule-based Autoscaling
Performance-based autoscaling in Azure monitors the average CPU usage and message queues. Azure can dynamically add Virtual Machine instances when CPU utilization exceeds a threshold set by an administrator, then turn the new Virtual Machine off when usage returns to the target threshold. Similarly, the number of messages in a queue can be monitored and Virtual Machines can be turned on or off from what’s known as an availability set.
If performance-based scaling is too reactive for your needs and the load on your servers is more predictable, schedule-based autoscaling can be used in a more proactive approach. For example, anticipated load based on events such as a go-live date, running of a TV commercial or a published registration date can be accommodated by configuring Azure to add resources on a targeted date and time. Used in conjunction with performance-based autoscaling, you can specify the number of additional Virtual Machines you think you will need, then let Azure add more if your target utilization thresholds are exceeded.
Azure’s benefits can be summarized by speed, scale and economics. The autoscaling features in Azure reflect all three:
- Speed – Setting up new environments can be done in minutes, with no manual intervention
- Scale – Size your environment appropriately based on intermittent, short-term or long-term needs
- Economics – Pay only for the resources used and then “turn it off” when they aren’t needed
If you want to learn more about Microsoft Azure’s autoscaling options, or are interested in a one day free on-site assessment to find out what it would take to move your environment to Azure, please join us for our upcoming Chicagoland Building and Migration Applications to Azure event or contact Ben Brock.
RECOMMENDED CLOUD POSTS
If you enjoyed this post on Microsoft Azure, read a few of SWC’s past blogs on the cloud:
Flexible and Scalable Web Hosting with Microsoft Azure
Five Questions to Ask Before Investing in a Cloud Service
What is the Private Cloud?
Why SWC And IT Departments Like Office 365
Ask SWC: How Can A Company Move To The Cloud With Privacy Needs?
Predictive Analytics Made Easy with Microsoft Azure Machine Learning
Microsoft Azure Machine Learning – A Data Tool for the Masses
Microsoft SharePoint: On Premises Or In The Cloud?
SWC’s 2013 Tech Trends Poll
What I Like About Microsoft System Center 2012
Is the Cloud Right for Your Mid-Market Business?
Let’s Build Some Cool Stuff Together
Watch our video below to learn more about SWC’s application development services.