Here’s my story about learning Google Cloud Platform (GCP) and passing the Google Cloud Engineer certification exam.
If you’re a software developer, sysadmin, or other kind of software engineer interested in working in Google Cloud with platform tools then this article is for you.
Read this guide because it is up-to-date and researched, and includes my actual personal exam guide, Mental Model diagrams, and a list of practical test-taking tips.
Rather than go over all the available options, I’ll share Google’s prescribed path and then share my personal study program, strategies and resources for passing the exam. As you will discover, you have quite a few different options.
I’m currently studying for my GCP Data and Machine Learning exam.
The ACE is no gimme
The GCP Cloud Engineer certification exam includes networking, systems administration, devops and development using a visual integrated development environment (IDE) and command line. So, there’s a lot to learn. Don’t complacently think you’re just going to slide into this.
Google advises starting with the ACE as a default. If you’re coming from AWS, maybe you will have an edge. But even experienced AWS Certified Professionals and seasoned cloud engineers may fail the exam. Because the ACE exam covers the entire fundamental knowledge-base, you could actually pass the Google Cloud Architect or other specialty certification exams and still fail the Google Associate Cloud Engineer exam.
I’m a software engineer and consultant, not an AWS expert. I passed my GCP ACE exam on the first attempt, but only after taking a deep dive into GCP and rescheduling the exam date more than once.
It takes discipline
You can’t do this full-time, you have to pace yourself. You need to be in shape. Your normal brain won’t handle this much input.
I uninstalled all my social media apps and put a sheet over the big-screen TV. I improved my diet and habits. I minimized caffeine and alcohol, made healthier eating decisions and made fresh vegetable juice daily. I took obscure yoga classes. I went to Pilates even more regularly. I meditated and visualized success.
GCP consists of a visual Console, including a menu where you access all the integrated Google Cloud Platform software programs and tools. You can do things like create and administer your network, control access and security, start servers, assemble components, deploy applications, build batch and streaming data pipelines, scale and load-balance, define firewall rules, work with data and machine learning tools, messaging, parallel processing, serve web apps, integrate mobile sdks, migrate, upload and store files, data and objects, and a lot more.
In some ways GCP is like an Internet Service Provider (ISP) with a cPanel for running application servers and databases, and other ways like a highly-featured visual software builder. Almost everything that can be done from the visual console can be done from the command line as well.
What you should study
First I’ll tell you how to do it right. I didn’t go this route, but since it is from Google I am listing the straightforward, recommended path from Google here first. Before you start, check out the resources on Google’s certification page, and focus on the skills from the Exam Guide.
1. Coursera GCP ACE prep class
This course is part of the Cloud Engineering with GCP professional Certificate multi-course program listed below. I did not do this class, but it is Google’s course specifically targeting ACE exam preparation.
Here is a link to the Cloud Engineering with GCP professional Certificate multi-course program. As mentioned, I didn’t take this program, but you should consider this because it is Google’s targeted offering.
Another Option: Udemy Cloud Guru GCP ACE exam prep course
Note: I took the Cloud Guru ACE prep class on Udemy along with another engineer, and we both passed the exam. It is very thorough and contains two practice exams at the end, however it is from a third party company. Take either class and you should be in good stead, but make sure you cover Google Kubernetes Engine (GKE). I explain this more further on in the article when I go over how I personally prepared for the exam.
2. GCP Essentials and Cloud Engineering QwikLab Quests
3. Udacity Scalable Microservices with Kubernetes
Note: Google doesn’t say you should take this class. Because the Cloud Guru course includes significant instruction on Google Kubernetes Engine (GKE), if you take the Udemy Cloud Guru GCP ACE prep course you don’t have to take the separate Kubernetes class.
Register for the exam through Google Webassessor and Kryterion Labs, and the company that does the actual exam in San Francisco is Learn-It. Do this so that you have something to shoot for. You can always reschedule, but there is a hefty rescheduling fee unless you give them more than 72 hours notice, in which case rescheduling is free.
What I studied
I actually didn’t take the Google Coursera class which is part of Google’s program — — as previously mentioned — — I took one on Udemy from Cloud Guru. And I also didn’t do the GCP Essentials Labs or Cloud Engineering QwikLab Quests, I chose labs ala carte. This is mainly because I didn’t properly know how Google’s courses and lab resources were organized or what was available and recommended by Google.
I took the Google Kubernetes (GKE) and Microservices class on Udacity.
I also did part of the Deep Learning and TensorFlow Course. You don’t have to do this now.
I completed the GCP ACE Udemy Cloud Guru course, which covers the topic thoroughly. The instructors tag team, so if one instructor doesn’t suit you, you might like the other.
Kubernetes is also covered in the Cloud Guru course — — there is an actual course-in-a-course for it. Google devotes a good part of the exam guide to GKE.
I didn’t read all of the outside reading and documentation fully while taking the Cloud Guru prep course. I read much of the documentation after identifying my knowledge gaps doing the hands-on labs. The Cloud Guru course included two practice exams at the end, which I failed with scores of 48%.
Four Strategies for Success
1. Identify your knowledge gaps
I used the failed practice exams from the Cloud Guru class to identify gaps in my knowledge, and collected GCP documentation links for each incorrect exam answer. The practice exams are an advantage of the course. Otherwise you can use the one from Google, but there’s only one and I would save that for when you need it, or else you run the risk of memorizing the questions.
2. Make a custom learning plan
I reorganized the exam guide according to my own mental model. First I copied the Google Exam Guide into a new Google Doc, then I copied the Google Cloud reading links from each resource section in the Cloud Guru course into the appropriate exam guide section. You can just google the web pages based on the exam guide, or you can read the documents listed in my personal study guide by following the link at the end of this article.
3. Track Progress with Color Coding
Once I was done with the course, I went through and read all the google cloud documentation links from my exam guide, and marked them to green. I colored the URLs red which I had not read yet, or had failed on the practice exam. I highlighted them in green after I covered the topic. I labeled the URLs I was uncertain about yellow, and changed the yellow highlighting to green after I had more clarity.
They say that GCP requires you to form a mental model. I drew around ten models covering different systems with arrows, flows, and dependencies. I recommend you draw diagrams of everything, or use mine as a starting point, and update them as you go along when you learn new info or discover that you are mistaken about something.
Below is the first Mental Model diagram I created. This Virtual Private Cloud (VPC) model sequence is somewhat arbitrary, but there are dependencies so you must do some things before others. The smiley faces in the Instance Group represent Compute virtual machines. See My Personal Study Guide link at the end of this article for my complete set of rough models.
About the Author
Software Engineer, Consulting CTO, Blogger
Blake Rogers is a Principal Engineer at Telegraph Hill Software, with 25 years software development experience in Silicon Valley. As senior software engineer and Interim CTO, he provides strategic technical leadership and hands-on development.