12 for '21 Cloud Optimization Series - Number Six: CloudWatch with Autoscaling

Cloud spend is anticipated to continue its massive growth trajectory for the foreseeable future.  However, the spend does not have to match that same growth curve!  To help our Clients maintain sound Cloud Optimization programs, we've assembled a series of articles and tips called: 12 for '21 Cloud Optimization Series (note: request a download of all 12 here).

Our sixth chapter in the series: CloudWatch with Autoscaling

AWS CloudWatch (Azure Monitor, GCP Monitoring) enables you to collect consumption metrics and set up alarms to invoke actions.  These tools collect these metrics, allows you to visualize them and to define alarms with the objective of optimal utilization of your cloud platform.   

Defining an Alarm

If you were monitoring your Virtual Machines (‘VMs’), the key consumption metrics of interest would be CPU Utilization/Memory Utilization/Network In/Network Out.  How do you get notified when one of these resources gets scarce or is  no longer being used? You would have to define an alarm and specify three settings to enable CloudWatch or Monitor to evaluate when to change the alarm state:

  • Period is the length of time to evaluate the consumption metric. It is expressed in seconds or minutes. If you choose one minute as the period, the alarm evaluates the metric once per minute.
  • Evaluation Periods is the number of  most recent periods (e.g., hours, days, weeks)  or data points, to evaluate when determining alarm state.
  • Datapoints to Alarm is the number of data points within the Evaluation Periods that must be breaching to cause the alarm to go to the Alarm state.

Alarm Actions

You can specify what actions an alarm takes when it changes state. The most common type of alarm action is to notify one or more people by sending a message. Alarms based on VMs  metrics can also perform  actions, such as stopping, terminating, rebooting, or recovering an EC2 instance.

AutoScaling and CloudWatch

AWS Auto scaling (GCP AutoScaling, Azure AutoScale) is a cloud computing feature  to automatically scale cloud services (e.g., VMs), up or down, depending on a  particular set of consumption  metrics  (e.g., CPU utilization, memory). Auto-scaling ensures that resources are seamlessly increased during demand spikes and decreased during demand drops, enabling consistent performance for lower costs. Using a combination of  CloudWatch data and its alarms, you can set up specific rules/ policies  within an Auto Scaling Group. With such, you could automatically scale up or scale down your infrastructure with a set of defined consumption metrics.  Between CloudWatch and AutoScaling, you not only save cloud money but also valuable time of your CloudOps team.

Contact us to learn more about this or any of our other Cloud cost saving strategies.

About NET(net)

Founded in 2002, NET(net) is the world’s leading IT Investment Optimization firm, helping clients find, get and keep more economic and strategic value. With over 2,500 clients around the world in nearly all industries and geographies, and with the experience of over 25,000 field engagements with over 250 technology suppliers in XaaS, Cloud, Hardware, Software, Services, Healthcare, Outsourcing, Infrastructure, Telecommunications, and other areas of IT spend, resulting in incremental client captured value in excess of $250 billion since 2002. NET(net) has the expertise you need, the experience you want, and the performance you demand. Contact us today at info@netnetweb.com, visit us online at www.netnetweb.com, or call us at +1-866-2-NET-net to see if we can help you capture more value in your IT investments, agreements, and relationships.

NET(net)’s Website/Blogs/Articles and other content is subject to NET(net)’s legal terms offered for general information purposes only, and while NET(net) may offer views and opinions regarding the subject matter, such views and opinions are not intended to malign or disparage any other company or other individual or group.

Read similar posts below