4 KPIs for Measuring Cloud Efficiency – Part 2

This is the second post on measuring cloud efficiency. In our previous post, we spoke about the first two KPIs- On Demand Coverage and Commitment Utilization.
There are two more indicators that can help us measure cloud efficiency.

KPI #3: Avg Hourly Cost

This is basically a way of combining a few different parameters; improving each one of the parameters will lead to a decrease in the hourly cost.
Obviously, the indicator is relevant for each service whose cost is measured while running, but users generally measure for one of the following 5 services:

  • EC2
  • RDS
  • Redshift
  • Open Search
  • ElasticCache

The Average Hourly Cost KPI is affected by the following parameters:

  • Machine size – the smaller the machine we use, the lower the hourly cost will be. Obviously, sometimes a strong machine is essential, making improvement irrelevant to this measurement, but in many cases, you can choose a less powerful machine at less cost (essentially ensuring right sizing) and consequently save money and reduce the average hourly computing cost.
  • Form of payment – running on On-Demand / Spot / Reserved Instance / Saving Plans. When using a cheaper payment option, such as Spot or machines covered by a commitment to your cloud provider, the hourly cost decreases, improving this entire metric.

The average hourly computing cost calculates a number of different parameters and therefore improving hourly computing costs usually indicates its improved efficiency in the cloud.

KPI #4: Percentage Waste

This indicator is more complicated to measure since you first need to define what you consider waste. In general, you can roughly divide the types of waste into three categories:

  • Idle resource
  • Resource that can get Right Sizing
  • Opportunity for commitment purchase

There are of course many services available from your cloud provider, and most of these services have resources that are not being used, or they use a more powerful machine than necessary. Some of them also have the option of committing to a certain amount of usage for the cloud provider to provide a discount.

The percentage waste KPI is built on a daily/weekly/monthly scan of cloud usage and subsequently finding usage reduction options. Then you calculate what percentage out of the total use is inefficient based on one of those 3 groups.

There are systems such as Pileus, that know how to perform these scans for dozens of Saving Opportunities automatically and allow a sustained percentage waste calculation. This enables you to measure whether there has been any improvement in the general waste out of all the cloud costs, or in the waste of a specific kind relative to the overall costs of said service.

Some common examples include the percentage of EBS Unattached out of the entire EBS costs, or the percentage of Old Snapshot out of the entire Snapshot costs, etc.

Conclusion

In conclusion, the ability to define the KPIs that will allow you to measure your cloud efficiency becomes critical in many organizations. This allows constant cost monitoring and the option for assessing what percentage of the costs are justified, and where there is room for improvement.

There are of course additional KPIs, but these four are doubtlessly the most significant indicators of your cloud account efficiency,