Auto-scaling (or Autoscale) is the process of allocating resources to match performance requirements and help minimize costs. Typically, this is an automated process that is set by policies or rules so you don’t have to continually monitor the performance and it can manage itself. As users demand more resources, the system is able to maintain the desired level based on the policies and as demand decreases, it can de-allocate those resources to help reduce costs.
Ways to Scale
Vertical Scaling: This type of scaling refers to changing the capacity of a resource and is not done as often. Vertical scaling also may require the system to be temporarily unavailable while it is being redeployed.
Horizontal Scaling: This type of scaling refers to adding or removing instances of a resources and is the more common of the two. Additional resources can be provisioned without interruption of current resources and they can be easily shut down without impacting other resources
While autoscale in Microsoft Azure can be a huge benefit, ensuring you take the time to create a strategy and implement it properly, will provide you the best results. Some things to keep in mind as you are planning out your strategy are that you will need to know what criteria you want to use, aggregation type, threshold, and what will trigger the rule.
Here is a breakdown of each category:
Criteria Metrics
• Percentage CPU
• Network In
• Network Out
• Disk Read Bytes
• Disk Write Bytes
• Disk Read Operations/Sec
• Disk Write Operations/Sec
• CPU Credits Remaining
• CPU Credits Consumed
Aggregation Type
• Average
• Minimum
• Maximum
• Total
• Last
• Count
Threshold Operator
• Greater than
• Greater than or equal to
• Less than
• Less than or equal to
• Equal to
• Not equal to
Triggers
• Increase count by
• Increase percent by
• Increase count to
• Decrease count to
• Decrease percent by
• Decrease count by
Check out some best practices for auto-scaling with MyCloudIT to see how we’ve helped simplify the process.
Tags: Cost Optimization