What Are the Advantages of Google Compute Engine?

Google Cloud Platform (GCP) is not yet a household name in the public cloud marketplace as is Amazon Web Services (AWS) or Microsoft Azure, but that’s changing rapidly. While the GCP still lags behind AWS and Azure regarding market share, it is catching up fast.

Many notable organizations, including HSBC, Snapchat, Philips, and Sony Music, are currently using GCP. Google has a strong value proposition in the public cloud marketplace because of its powerful Google Compute Engine (hereafter just called Compute Engine). Organizations can use the Compute Engine to run fault-tolerant, high-performance, scalable Virtual Machines (VMs) on-demand.

What is Google Compute Engine?

The Google Compute Engine is an Infrastructure-as-a-Service (IaaS) offering within the GCP. Rather than procuring and managing server hardware and its accompanying resources, you can leverage the Compute Engine and run powerful VMs on Google’s public cloud.

Below are some features of the Compute Engine:

Machine Types

The Compute Engine uses a Kernel-based Virtual Machine (KVM) as its hypervisor. You can use Compute Engine to launch guest images running Linux-based and Microsoft Windows Server OS. Compute Engine provides two methods for launching VMs: the pre-configured and the custom approach.

Under the pre-configured method, users leverage pre-configured templates to set up their VMs. There are four categories of VMs that range in purpose within the pre-configured approach:

Standard VMs. These are balanced between computational power and memory and are suited for most workload requirements.
High-memory VMs. They are optimized for memory-intensive tasks that require access to non-disk storage quickly.
High-CPU VMs. They are optimized for intensive computational workloads.
Shared-core VMs. These VMs timeshare a physical core and are cost-effective for running small and non-resource-demanding applications.
Accelerator-optimized machine types. These machines have high performance and are suitable for parallel computing workloads such as high-performance computing and machine learning.

You can also customize your machine type manually. Under this approach, you need to select the number of virtual CPUs (vCPUs) and memory, provided you are within Google’s set limits.

Images

An image contains the OS and the root file system that users leverage to run VM instances. Google provides two main types of images: public and custom images. Public images contain both open-source and proprietary options.

If you’re just starting or exploring GCP, you can use the open-source CentOS and Debian images that Google provides as standard images. You can also use the proprietary images—mainly Red Hat Enterprise Linux (RHEL), Microsoft Windows Server, and now Windows client —for more cloud options, but at an extra cost.

While public images are a great starting point, they may not offer customized needs for your organization. Custom images are great for customized solutions such as production-ready environments. This is because they provide the software you need and have the necessary scripts that developers can use to get started without the intervention of IT administrators.

Storage

The Compute Engine provides three storage options for VM instances: persistent disks, Filestore, Local SSD, and cloud storage.

Persistent disks are block-oriented systems and, as the name suggests, build data persistence whenever VMs start, stop, or terminate. The Compute Engine uses Small Computer System Interface (SCSI) to attach persistent disks. Persistent disks can be zonal or regional. Zonal persistent disks are efficient, reliable block storage, easy to use, and have high performance and reliability. Regional persistent disks are replicated in two zones.
Filestore is a fully managed storage option that provides network file storage. If you are implementing a system with multiple parallel services, you can use Filestore to access the files from the same disk over the network. On the other hand, cloud storage is redundant storage that you can mount onto a VM’s file system, just like persistent disks.
Local SSD, as the name suggests, is attached to the VM instance’s server physically. With high throughput and lower latency, the data stored in them persists until the instance stops or is deleted.
Unlike persistent disks, which are file-based and can serve as a root drive for the VM, cloud storage is object-based and cannot serve as a root in the file system.

Practical Use Cases for Adopting Compute Engine

Below are some of the applications of Compute Engine:

VM migration. Compute Engine has tools to fast-track migration from on-premise or other clouds to GCP. Starting with the public cloud, you can leverage these tools to seamlessly transfer existing applications from your data center, AWS, or Azure to GCP.

Genomics data processing. Processing genomic data is computationally-intensive because the information is enormous with vast sets of sequencing. With the Compute Engine’s potential, you can process such massive data sets. The platform is flexible and scalable when it comes to processing genomic sequences.

Bring Your Own Licence (BYOL) images. Compute Engine can help you run Windows apps in GCP by bringing their licenses to the platform as either license-included images or sole-tenant nodes. You can flexibly optimize your license and promote your bottom line when you transition to GCP.

Google Compute Engine Advantages

Compute Engine offers an ideal solution regarding throughput, stability, pricing, backups, and security when it comes to the public cloud marketplace.

1. Throughput

Compute Engine’s network Input/Output across regions is much faster than AWS’s. Google’s global network infrastructure—the backbone of the Compute Engine—is superior to AWS, which uses the public internet. As of Q1 2020, Google has 22 regions and 61 zones for its Compute Engine infrastructure. Also, Google is investing billions of US dollars in cementing its footprint in cloud computing.

This provides 100% uptime via transparent maintenance compared to AWS and Azure. You can set up multiple cloud scenarios, including synchronous database replication between the regions.

2. Efficient Block Storage

Compute Engine’s persistent disks can support up to 257 TB of storage. This is more than ten times higher than what Amazon Elastic Block Storage (EBS) can accommodate (currently, the maximum is 16 TB). As such, Compute Engine is best suited for those organizations that want more scalable storage options.

3. Stability

Compared to AWS, the Compute Engine offers more stable services because of its ability to provide live migration of VMs between the hosts. This means organizations can run 24 hours a day, seven days a week, and 365 days a year without downtime or other performance hindrance.

4. Pricing

Within the GCP ecosystem, you pay only for the computing time you have consumed. The Compute Engine uses the per-second billing plan, as opposed to AWS, which is per-hour based. You are also entitled to attractive discounts for long-running workloads on the Compute Engine. While Azure also provides discounted rates, you get only a 5% discount for a whole year’s pre-payment, compared to Compute Engine’s 30% discount for a month’s pre-payment.

5. Backups

GCP has a robust, inbuilt, redundant backup system. The Compute Engine uses this system for its flagship products, such as the Search Engine and Gmail.

6. Security

It has been more than 20 years since Google launched. When you choose GCP, you get the security benefits that Google has developed over the years to secure its robust products, such as the Search Engine and Gmail.

Parallels RAS on Google Compute Engine

Parallels® Remote Application Server (RAS) is a cost-effective solution and an industry leader for virtual desktop infrastructure (VDI) solutions, and it’s simple to deploy and maintain. It supports major hypervisors, including Microsoft Hyper-V and VMware ESXi. Parallels RAS also supports Google Compute Engine as a VDI solution.

When deployed on the Compute Engine, Parallels RAS allows companies to provision and scale their VDI workloads directly. Parallels RAS has cloud automation features. When combined with Compute Engine’s inbuilt VM templates and configuration wizards, it is a perfect fit for organizations that want to deliver VDI solutions faster and more reliably.

Most of all, Parallels RAS is simple to install on Google Compute. You can easily set up a Parallels RAS environment on the Compute Engine and start to deliver and manage your VDI workloads in minutes!

Realize the benefits of VDI on Compute Engine!

Download the Trial

Growth Opportunities for MSPs, ISVs, VARs and SIs in the Post-Pandemic Era

Read our white paper

The Road from VAR to MSP: How to Successfully Transition from One-Off to Recurring Revenue

Read our white paper