Running cost effective Virtual Machines in the Cloud

PublishedFebruary 28, 2023

I'm technologist in love with almost all things tech from my daily job in the Cloud to my Master's in Cybersecurity and the journey all along.

Comments

Join the discussion

No comments yet. Be the first to comment.

More from this blog

How Large Language Models Learn, Part 2

In Part 1 of this series we explored the friendly foundations of machine learning: Classification Regression Clustering Neural networks Training Testing Overfitting Decision trees Reinforceme

Jun 26, 202617 min read

A Friendly Guide on How Large Language Models Learn

Machine Learning with Trufa and Paula: A Friendly Guide to How Models Learn

May 14, 202613 min read

A Friendly Guide on How Large Language Models Learn

My hands on with Copilot Studio

The Microsoft Copilot Ecosystem

May 8, 202616 min read

Vibecoding in practice - An online Dictionary en español

I love words, I compile words and always search what they mean. My native language is español and to my surprise there are not that many on line. México where I'm from has no official online language

Mar 12, 20264 min read

Vibecoding in practice - An online Dictionary en español

The Power BI Supercharge: How Fabric, OneLake, and DirectLake Change the Game

Hey everyone, Roberto here and there and everywhere! We've all worked with data, we've all felt the pain of waiting. When you build a beautiful Power BI report, but the data is from yesterday. You have to wait for the nightly refresh to see the lates...

Feb 7, 20266 min read

The Power BI Supercharge: How Fabric, OneLake, and DirectLake Change the Game

U

Understand. Build. Conquer the Cloud

73 posts

No time for a novel? Here are my my Cloud Architect field notes: Distilling my complex cloud adventures into digestible TL;DRs.

The logic makes sense, Cloud Providers have excess capacity and instead of having it there idle generating heat and getting depreciated better use it and sell at cheap(er) price.

Customers gain because they pay less and can use them for lower-priority tasks or Hyper performance batch processing, any service that doesn't require continuous running.

Preemptive Virtual Machines are excess Compute Engine capacity and their availability varies. A VM of this type can be terminated by the Cloud vendor at any time if it requires access to the underlying resources for other tasks.

In Google Cloud

The Compute Engine always terminates instances of this type after they run 24 hours.
They might not always be available for provisioning
These types of instances are not covered by the Google Cloud SLA
These types of instances are not available in the Free-Tier

In Azure

There's a service in the preview for Scalesets using Low-priority VMs:

The amount of available unutilized capacity can vary based on size, region, time of day, and more. Azure will allocate the VMs if there is capacity available, but there is no SLA for these VMs. A low-priority scale set is deployed in a single fault domain and offers no high availability guarantee.

In AWS

Amazon can interrupt your Spot Instance when the Spot price exceeds your maximum price, when the demand for Spot Instances rises, or when the supply of Spot Instances decreases. When Amazon interrupts a Spot Instance, it provides an interruption notice, which gives the instance a two-minute warning before Amazon interrupts it.

Roberto

Sources

#azure

Command Palette

Comments

More from this blog

In Google Cloud

In Azure

In AWS