Google Compute Engine (GCE) is an Infrastructure-as-a-Service (IaaS) part of the Google Cloud Platform (GCP). With Cloudways, you can build and deploy high-performance and scalable applications on GCP with simplicity and freedom. Cloudways Managed Hosting Platform + Google’s infrastructure — is a formula that guarantees unmatched performance, reliability, and choice with 24/7 support that acts as your extended team.

GCE Servers Family

To deliver you with the best and excellent hosting experience, the GCE server lineup is divided into three main categories:

  1. Standard Series

  2. Compute-optimized a.k.a. CO series

  3. General Purpose High-CPU a.k.a HC series

Each family is designed for specific workload types. Choosing the right server for your workload is a crucial part of your application’s success. Let’s dive in a bit deep to uncover what these new instances offer.

1. Standard Series

Standard series is the first generation general-purpose series of Google Compute Engine (GCE). Standard series machines are available on Intel Skylake, Broadwell, Haswell, Sandy Bridge, and Ivy Bridge CPU platforms based on availability in different regions. These machines have local SSDs having higher throughput and lower latency than standard disks.

Standard series also have one instance called “Small” which has 1.70 GB RAM and 1 shared vCPU for short periods of bursting. Please note that the “Small” instance does not have a local SSD.

Server Name

Processor

Memory

Caremark Benchmarks

Small

1 shared vCPU

1.70 GB

10,657

Medium

1 vCPU

3.75 GB

20,090

Large

2 vCPU

7.50 GB

26,532

XL

4 vCPU

15 GB

54,678

2XL

8 vCPU

30 GB

105,383

4XL

16 vCPU

60 GB

221,834

8XL

32 vCPU

120 GB

440,755

2. Compute-optimized (CO) Series

Compute-optimized family instances are powered by the blazing-fast latest generation Intel Cascade Lake processors and robust local SSD drives. Local SSDs have higher throughput and lower latency than standard disks. The processor offers a base frequency of 3.1 GHz, and up to 3.8 GHz sustained all-core turbo. CO series is designed to achieve ultra-high performance for compute-intensive workloads as it offers the highest performance per core.

The 3.8 GHz CPU offers faster content delivery due to more efficient processes and script executions. Besides powerful spec sheets and additional computing power, Cloudways Thunder stack and advanced caching techniques optimize these instances on top of it to maximize the performance so you can make the most out of it. The following servers are available under this family:

Server Name

Processor

Memory

Coremark Benchmarks

CO XL

4 vCPU

16 GB

77,310

CO 2XL

8 vCPU

32 GB

148,689

CO 4XL

16 vCPU

64 GB

313,768

CO 7XL

30 vCPU

120 GB

571,147

CO 14XL

60 vCPU

240 GB

1,142,234

3. General Purpose High-CPU (HC) Series

General Purpose High-CPU family instances are designed to offer the best price-performance ratio for the different workloads. It is ideal for those workloads and tasks that require a moderate increase of vCPUs relative to memory. These instances run on the latest generation Intel Cascade Lake processors with 2.8 GHz base frequency, and up to 3.4 GHz sustained all-core turbo. It also offers robust local SSD drives. Local SSD drives have higher throughput and lower latency than standard disks.

Operations requiring higher clock frequency can take advantage of such instances as they offer higher per-thread performance while benefiting from Cloudways Thunder stack and advanced caching techniques, which optimize these instances on top of it. The following servers are available under this family:

Server Name

Processor

Memory

Coremark Benchmarks

HC Large

2 vCPU

2 GB

33,486

HC XL

4 vCPU

4 GB

68,980

HC 2XL

8 vCPU

8 GB

132,614

HC 4XL

16 vCPU

16 GB

279,395

HC 8XL

32 vCPU

32 GB

558,827

HC 16XL

64 vCPU

64 GB

1,110,895

Availability

This GCE lineup is subjected to availability from the Google Cloud Platform (GCP). GCP has 14+ data centers across the globe. Please note that Mumbai, India data center does not offer General Purpose High-CPU (HC) instances.

Please note that if you are scaling up in a region where your desired server is unavailable, then you can clone your server to a different region to get that machine.

Benchmarks

We have benchmarked a few GCE instances to compare the performance. This benchmark includes a Page Response comparison of each server from the following series.

  • Standard Series — XL, 2XL, 4XL, 8 XL

  • Compute Optimized (CO) — CO XL, CO 2XL, CO 4XL, CO 7XL, CO 14XL

  • General Purpose High-CPU — HC XL, HC 2XL, HC 4XL, HC 8 XL, HC 16XL

Tip

Page Response evaluates the number of responses the server can handle under a sustained load from the users

Test Environment Conditions

All the benchmarks were carried out using the same testing conditions for each machine.

Server Region

USA – Iowa

Server Application

WordPress

Test Duration

1 Minute

Concurrent Clients

1-250 Client

Load Test Algorithm

Maintain Load

Page Response Comparison (Varnish Cache Disabled)

The first comparison was between the chosen machines keeping Varnish cache disabled. It is being observed that after CO 7XL and HC 8XL shown in the graph as 7-8XL (x-axis), the rise in request count (y-axis) is not significant considering the previous trends. However, the Compute Optimized (CO) machines were able to handle a higher number of requests and performed well. On the other hand, the Standard Series machines after 4 XL did not show greater response counts.

Page Response Comparison (Coremark Score)

The second test was to compare Coremark scores among our chosen machines. Coremark scores are impressive, and all the machines show consistent performance. On a performance level, both CO and HC families compare well. Yet, we need to keep into perspective the amount of memory both machine types are equipped with, so if your application demands a good amount of memory, CO machines will handle the job better.

Launching a New Application On a GCE Server

GCE Server — Accessing the Cloudways Platform

Log in to your Cloudways Platform using your credentials.

  1. From the top menu bar, open Servers.

  2. Then, click Add Server.

GCE Server — Launching a New Server

  1. Select the application you want to deploy.

  2. Name your application.

  3. Next, type the name of your server.

  4. Select your project.

  5. Select the Google Cloud Platform (GCP) as an infrastructure provider.

  6. Next, choose your desired server size.

  7. Choose the amount of your expected bandwidth consumption per month.

Tip

It is recommended to deploy your server on the data center closest to your clients to minimize the network latency.

8. Scroll down and select the storage size for application web files.

9. Next, choose the storage size for the database.

10. Choose your server location from the list.

11. Finally, click Launch Now.

That’s it! Your new GCE Instance will be deployed in a few minutes.

Tip

Once your server and application(s) are created, then don’t forget to follow the take site live checklist to take your website online.

Upgrade an Existing Server to a New GCE Server

If you use Google Cloud Platform already, you can easily scale up to a new GCE family using a vertical scaling feature. Else, if you use any other cloud partner and want to switch, you can clone your server.

Moving from the Existing GCE Server

You can scale up your current GCE instance to the latest generation servers using the following steps:

Important

  • We recommend taking on-demand server backup before doing any scaling operation on your server; this will ensure you have a good restore point if you need to restore your applications’ previous versions.

  • The scaling operation will take your server offline while resources are being allocated to your server. The process should be lean and swift, yet it is good practice to scale during off-peak hours to minimize your website’s downtime effect.

Log in to your Cloudways Platform using your credentials.

  1. From the top menu bar, open Servers.

  2. Then, choose your desired server to scale up.

3. Under Server Management, select Vertical Scaling.

4. Under server size, hover over different server sizes to get more information. Once decided, you can select that server size.

5. Finally, click Scale Now.

Important

Please note that if you are scaling up in a region where your desired server is unavailable, then you can clone your server to a different region to get that machine.

6. You will be prompted with some warnings about the scaling process. Once you are ready, you can click Proceed.

It may take a few minutes to get your server up and running again with new specifications.

Swapping Your Current Server with a GCE Server

Upgrading from other cloud providers to a new GCE server is possible by cloning the server.

Log in to your Cloudways Platform using your email address and password.

  1. From the top menu bar, open Servers.

  2. Then, select your desired server.

  3. Click and click Clone Server.

4. You will be prompted for the cloning process. Click Yes.

5. Select the Google Cloud Platform (GCP) as your infrastructure provider.

6. Next, select your desired server size.

7. Choose the storage size for application web files.

8. Next, choose the storage size for the database.

9. Select your server location.

10. Finally, click the Clone Server.

Your new instance will be deployed in a few minutes. Please bear in mind that the cloning process only creates an identical copy of your website contents. It does not merge previously applied configurations or settings through the Platform. Once the cloning process is completed, you will need to follow the take site live checklist and update your DNS records to get your website online.

That’s it! We hope this article was helpful. If you need any help, then feel free to search your query on Cloudways Support Center or contact us via chat (Need a Hand > Send us a Message). Alternatively, you can also create a support ticket.

Did this answer your question?