In this article, I’d like to share my journey to getting GCP Certified while already having several AWS certifications (AWS Solutions Architect Pro, AWS DevOps Pro and AWS Security Specialty).

I will also highlight the main differences between AWS and GCP certification programs and exams.

Sometime around March 2023, after getting my AWS Solutions Architect Professional and AWS Security Specialty certifications, I started thinking about what to learn next. Given that another team in my domain was fully operating on GCP, my manager suggested that I dive deeper into Google Cloud. And what better way to immerse myself in technology than by preparing for certification? 

I already had some experience with GCP, mainly with BigQuery and Cloud Pub/Sub so the journey to becoming GCP certified was expected to be exciting and full of discoveries.

About GCP Certification program and how it compares to AWS

• There is one entry-level certification: Digital Cloud Leader, While AWS offers the Cloud Practitioner in the category.
• There is only one associate-level certification: Cloud Engineer. Just for comparison AWS offers three associate-level certifications here, with a fourth expected to arrive in March 2024.
• GCP offers nine professional-level certifications, including Cloud Architect. In contrast, AWS offers two professional-level certifications here and six Specialty certifications. The Specialty certifications are considered Pro but focus on areas with a narrower scope.

???? Full description of GCP certifications – https://cloud.google.com/learn/certification

???? Full description of AWS certifications – https://aws.amazon.com/certification/

Key differences between AWS and GCP certification programs:

⏱ Validity Period: all AWS certifications are valid for three years, compared to only two years for GCP professional certifications.

???? Testing Vendors: AWS partners with Pearson Vue vendor, while GCP uses Kryterion as its testing vendor. This leads to differences in testing software and the availability of onsite test centers. Fortunately,  in Berlin there is enough selection of both.

✏️ Exam Feedback: Both GCP and AWS exams follow a PASS or FAIL grading system. However AWS provides a detailed exam report with the final score and performance in each section while GCP doesn’t provide anything.

???? Exam Content: 

• The GCP Cloud Architect exam includes questions based on predefined case studies, which are available to candidates both before and during the exam. Learn more about case studies
• The AWS SysOps Administrator exam used to include three labs, offering a hands-on experience that many found to be a more effective assessment of AWS skills. Despite their popularity, AWS has temporarily removed these labs from the SysOps exam.

Preparations

So the choice of learning path was quite simple: Cloud Engineer, then Cloud Architect Pro, maybe Cloud Security Engineer in the future (spoiler alert: I still haven’t started with this one). For preparation, I used Udemy courses by 28Minutes on Cloud Engineer and Cloud Architect along with practice exams. Big thanks to Delivery Hero for providing unlimited Udemy access. ????

The preparation took around three weeks for the Cloud Engineer exam and an additional two weeks for the Cloud Architect one. The Cloud Architect curriculum builds upon the Cloud Engineer one. In my experience, there is more than a 50% overlap between these curriculums. 

The exams

Due to past issues with online exams, I always choose onsite options to avoid any unexpected problems. For both my exams, I chose PC-COLLEGE Berlin near Anhalter Bahnhof as my test center. I passed the Cloud Engineer exam on 7th April, and the Cloud Architect on 22th April.

Surprisingly, I found the Cloud Architect exam to be simpler than the Cloud Engineer exam. This is because the Cloud Engineer exam requires hands-on experience with Google Cloud CLI tools, whereas the Cloud Architect exam is more theoretical covering broader topics like CI/CD and Agile methodologies etc. 

My personal take on AWS vs GCP differences

And in conclusion, I’d like to point out some most interesting differences between GCP and AWS from my personal perspective:

• IAM Differences: Transitioning from AWS IAM to GCP IAM was challenging:
  • In AWS, an IAM Role serves a similar function to a GCP Service Account, while a GCP Role corresponds to an AWS IAM Policy.
  • GCP supports easy cross-account access by simply adding a principal to a project, while AWS requires to provision of an IAM Role and set trust relationships; 
  • Also, AWS supports some handy IAM conditions like aws: PrincipalOrgID (which allows you to share, for example, a bucket with the entire organization) – a feature I haven’t found in GCP. Plus, there are plenty of other smaller differences.
• GCP consolidates messaging into a single service: Cloud Pub/Sub, which basically does the same thing as classic AWS SNS + SQS and even acts as Kinesis Data Streams. The pricing model for these services is totally different, hence cost optimization techniques are different too.
• AWS Security Groups are applied at the Elastic Network Interface (ENI) level, allowing attachment to various sources such as load balancers, compute service instances, RDS instances etc. Whereas the GCP Cloud Firewall operates at the subnet level. This setup means that what’s straightforward in AWS – like setting up a security group to allow traffic to a specific load balancer only from Cloudflare IPs, free of charge- isn’t as simple in GCP. You could use Cloud Armor for a similar purpose, but it might become expensive at scale.
• Switching between projects in the GCP console is much simpler than compared to the frequent need to log in again between different AWS accounts.
• Serverless compute services: AWS heavily promotes AWS Lambda, embracing a serverless-first strategy that positions Lambda as a versatile tool capable of supporting nearly any workload—a focus clearly reflected in their curriculums and exams. In contrast, Google Cloud functions (both v1 and v2) seem to receive less attention in blogs and announcements, often appearing overshadowed by Google Kubernetes Engine GKE.
• Key features of GCP Compute Engine that are missing in AWS:
  • Sustained Use Discounts: Get a discount on your Google Compute Engine (GCE) instances by simply keeping them running continuously!
  • Live Migration: Move a VM to another host without interruption. In contrast,  AWS EC2 instances require a stop-and-start for such tasks.
  • Custom GCE Machine Types: No limitations of predefined cloud provider instance types and can create your own.
• Many Google Cloud API need to be enabled before they can be used.

Conclusion

I’d like to say that the journey to becoming GCP certified was definitely worth the effort and has positively impacted my daily job. And of course, getting a certificate is extremely self-satisfying.