Run Fault Tolerant Non Critical Workloads with Preemptible VMs | Scribe

Run Fault Tolerant Non Critical Workloads with Preemptible VMs

  • Hafeez Baig |
  • 9 steps |
  • 20 seconds
  • Google CloudGoogle Cloud
1
Sign in into the **Google Cloud Platform**
2
Type "**Compute Engine**" in the search bar and click on the **Compute Engine** option
information ordinal icon
**What is Compute Engine?** Google Compute Engine (GCE) is an Infrastructure-as-a-Service (IaaS) offering from Google Cloud Platform (GCP) that allows users to run virtual machines (VMs) on Google's infrastructure. It provides scalable and flexible computing resources, high-performance persistent storage, advanced networking capabilities, and integration with other GCP services, making it suitable for a variety of workloads including web hosting, data processing, and high-performance computing.
3
**VM instances** wizard will open, click on the **CREATE INSTANCE** button on the top left to create a new instance.
information ordinal icon
**What is an Instance?**\ \ An instance is a virtual machine (VM) running on Google Cloud Platform's Compute Engine, providing the necessary resources and environment to run applications and workloads.
4
The **Create an instance** wizard will open. Enter the Name as **my-first-vm**.
5
Navigate to the **Advanced** section by selecting it from the left-hand sidebar.
6
In the **Advanced** section, scroll to the **VM provisioning model** section and select **Spot** option from the dropdown.
information ordinal icon
The Google Cloud Platform has rebranded **"preemptible VMs"** to **"Spot VMs."** Spot VMs offer the same cost-saving benefits and are designed for fault-tolerant workloads. The change emphasizes the use of these VMs for workloads that can tolerate interruptions, providing up to 91% savings compared to standard VMs. For in-depth information about Spot VMs, please refer to the official Google Cloud documentation on [Spot VMs](https://cloud.google.com/compute/docs/instances/spot).
7
Scroll to the **On VM termination** section, here you can choose the options to **Stop** or **Delete** the instance on VM termination.
information ordinal icon
**What does the "Stop" option under "On VM termination" mean?** The "Stop" option under "On VM termination" in GCP means that when the VM instance is terminated, it will be stopped and can be restarted later, rather than being permanently deleted. This allows you to preserve the instance's state and data for future use.
8
Scroll to the **Host error timeout** section, here you can select the Host error timeout for your VM instance.
information ordinal icon
**What is Host error timeout?** Host error timeout in GCP refers to the duration a virtual machine (VM) waits for a host to recover from an error before taking further action, such as migrating the VM to another host or stopping it. This ensures that temporary host issues do not immediately disrupt the VM's operations.
9
Scroll the page and click on the **CREATE** button
information ordinal icon
Congratulations! on completing this lab and Understanding Run Fault Tolerant Non Critical Workloads with Preemptible VMs