IMG 20221029 075238 150x150

Relationship between Rightsizing resources, Carbon Footprint & Cloud expenses

published at 11-08-2022 by Vanessa Kantner

The technique of matching cloud instance types and sizes to the workload performance and capacity requirements at the lowest possible cost is known as Rightsizing.

The key compute utilization metrics to consider while rightsizing are vCPU, Memory, Network and Disk use (Virtual IT Resources). Usage patterns play a key role in rightsizing, to properly estimate rightsizing, we need to analyze cloud usage patterns.

What does Rightsizing mean?

Based on usage patterns of different resources, cost relationships can be analyzed and predicted. The following are the different usage patterns with their cost-related analytics

  1. Consistent load: can use `use-based discounts´ and `spend-based discounts´ for cost efficiency as the resources are constant over the period of time.
  2. Variable load but predictable: might use a mixture of Savings plan and Reserved instances to reduce cloud expenses.
  3. Variable load but not predictable: should be forecasted and monitored on a regular basis for constant rightsizing to improve efficiency. A low-cost savings plan can cover such loads.
  4. Temporary load: which has flexible start times can use Spot instances rather than on-demand resources.
  5. Special cases (test and dev environments): can be switched off during weekends to decrease the cloud expenditure.

Benefits of rightsizing:

These are the three major benefits we can obtain due to rightsizing

  1. Rightsizing decreases cloud expenses
  2. Rightsizing helps you to understand the workload needs better
  3. Rightsizing also helps you to manage your cloud infrastructure efficiently.

A simple 3-step procedure for rightsizing:

  1. Analyze: Analyze the utilization of your infrastructure.
  2. Check: Determine whether these are running efficiently and what actions you can take to improve efficiency.
  3. Action: Upgrading, Downgrading, and Terminating the infrastructure according to the need.
  4. Upgrading Resources - after significant analysis over a period, upgrading is recommended for workloads with high utilization.
  5. Downsizing or Downgrading Resources - Downsizing is recommended for Underutilized resources which can achieve the same performance even after downsizing to a certain extent.
  6. Deleting or terminating Idle Resources – Deleting resources which are not in use results in immediate cost savings. Such resources are also called Zombie assets.

Relationship between Cloud usage and Cloud expenses

The greater the cloud usage efficiency, the lesser will be the cloud expenses. So, in order to reduce cloud expenses, you must focus on cloud usage, as cloud usage and cloud expenses are directly proportional to each other.

In the cloud, you don’t pay for what you use, you pay for everything you run whether you use it or not

The following are the crucial rightsizing principles to reduce cloud usage

  1. Rightsize before performing a migration
  2. The best rightsizing starts from day 1
  3. Measure Twice, execute rightsizing Once
  4. Test once and perform multiple rightsizing
  5. Combine Reserved Instance and Savings Plans strategies with rightsizing to maximize savings

Therefore, any effort to reduce cloud usage reduces the proportional amount of costs.

Relationship between Rightsizing and Carbon footprint

As we all know cloud usage is directly proportional to the amount of carbon footprint, but efficient use of cloud resources will automatically decrease the amount of cloud carbon footprint generated (which will be less than usual)

Steps to reduce the carbon footprint:

  1. Rightsizing Resources
  2. Deletion of idle resources
  3. Switching off resources during idle hours
  4. Choosing the region which utilizes renewable energy

In the cloud, co2 footprint is not only released from the resources that are in use but co2 footprint is released from everything you run whether you use or not

Therefore, any inefficiency in cloud resource usage results in a larger carbon footprint.