Although one of the primary reasons for choosing Google Cloud Platform (GCP) to be your cloud services provider is cost, you should never miss an opportunity to optimize GCP spend. Billions of dollars are wasted each year on cloud costs despite there being dozens of tools and services on the market to reduce spend, increase performance, and enhance governance.
Google itself provides a number of tools to help optimize GCP spend. Its Stackdriver monitoring tool now integrates with Amazon Web Services and a number of other native and open source programs to help businesses operating in multi-cloud and hybrid cloud environments better manage their infrastructures. The cloud service provider also offers a rightsizing tool that extracts data from Stackdriver in order to recommend appropriate configurations for VM instances.
However, by Google´s own admission, both the Stackdriver monitoring tool and the rightsizing tool have their limitations. Stackdriver doesn´t work with Microsoft Azure or Alibaba Cloud, and is incompatible with several other key installations. Google´s rightsizing tool makes its recommendations based on a maximum of eight days of data, and is therefore not beneficial to businesses whose VM instances experience monthly peaks and troughs in demand.
The ParkMyCloud Way to Optimize GCP Spend
ParkMyCloud differs from Google´s optimizing tools in that it provides total visibility across all your accounts with all major cloud service providers. ParkMyCloud integrates seamlessly with the CloudHealth platform to monitor and analyze costs, usage, security and performance, and makes rightsizing recommendations to optimize GCP spend based on user-defined time periods, rather than being limited to eight days CPU and memory utilization.
Using ParkMyCloud´s single-pane dashboard, administrators can easily identify and terminate unused “zombie” assets. They can manage role-based access controls to enhance accountability and security, and enforce governance through policy-driven automation. However, the most cost-effective use of ParkMyCloud is to schedule on/times for non-production VM instances – typically those used for development, staging and testing.
Scheduling On/Off Times to Reduce GCP Spend
In order to work out exactly how effective ParkMyCloud is at reducing spend on GCP, it has to be assumed that non-production VM instances would not be included in a committed use discount program; but, if left running at all times, would qualify for a sustained use discount. Therefore, taking an n1-standard-8 VM instance as an example (located in Northern Virginia), we know that – after applying the sustained use discount – the cost of running the VM instance for 30 days would be $215.71. (*)
If you were to apply a schedule that switched the VM instance on between 9.00 a.m. and 5.00p.m. Monday to Friday, the VM instance would be running for 168 hours each month (based on twenty-one working days in the month). Although the VM instance would no longer qualify for a sustained use discount, the cost of running it for the month would be $71.90 – a saving of 66 percent. However, reducing spend and optimizing spend on GCP are not the same thing.
(*) $0.4280 x 720 hours – minus 20 percent discount after 180 hours, 40 percent discount after 360 hours, and 60 percent discount after 540 hours
Optimizing GCP Spend for Your Particular Usage
For some businesses, applying a standard 9.00 a.m. to 5.00 p.m. schedule will suit their needs perfectly and optimize their GCP spend as a result. However for other businesses, demand for non-production VM instances can fluctuate throughout the day, or their development teams may work in irregular patterns – making the application of a regular schedule impractical. In both scenarios, ParkMyCloud has a solution.
For businesses with fluctuating demand, ParkMyCloud offers “SmartParking” recommendations. These are based on historic CPU utilization and can be applied with the click of a mouse. If your development teams work in irregular patterns, you can “Always Park” non-production VM instances and interrupt the schedule when access to the resources is required. Using these two options, you can reduce GCP spend by more than the 66 percent quoted above and optimize GCP spend for your particular usage.
Find Out More about How to Optimize GCP Spend the ParkMyCloud Way
If your business is deploying resources on Google Cloud Platform either in a single cloud environment or a multi-cloud environment, and you would like to find out more about how to optimize GCO spend the ParkMyCloud way, do not hesitate to contact us. Our team will be happy to answer your questions, arrange a free demo of ParkMyCloud in action, or invite you to take a free trial of our lightweight SaaS app so you can evaluate ParkMyCloud in your own environment.