Motivation

I think there are many reasons to decide and take on a professional certification: learning, enhancing your resumé, being strongly advised by your management for a given position, etc.

In my case, it was rather curiosity: my employer announced this summer they would pay for a Google certification, there was a choice of 3 as far as I remember: Professional Cloud Architect, Security Engineer, and Data Engineer.

I have a decent AWS and Terraform experience (3+ years of professional experience), same for Kubernetes, and I played around with Google Cloud several times. Could be interesting and should be a walk in the park, right? Well… so I thought 😖 ..

Going through all the certification program stages

After I registered for the program (July 16th), I was accepted into a cohort (August 19th) for a Stage 1 (few labs to complete in 7 days) that eventually gave access to the main event: Stage 2 where lessons would be given every week (until November) with an instructor, and at least 5 badges (out of 6) should be earned before asking a voucher for the 2 hours long multiple choices proctored exam (latest date December 30th)

So it’s a 6-month long experience, lessons are recorded in case you can’t make room for the Stage-2 2 hours long 10AM to 12PM weekly Tuesday class (I could not attend any of them 😩), and the badges can be earned when you complete a series of labs, some of them being almost 8 hours long!

The classes

Well, I skipped them, so I can’t really describe them 😅 ; but one thing you should know: they’re about preparing the exam, not hands on Google Cloud services; for this you have the self paced labs and their badges

The labs

For the most part, they’re pretty interesting: you receive Cloud Credits that you can spend on labs; each lab will then provision a temporary account (for 30mn to 2 hours), where you’ll discover Google Cloud Services through exercises using the Google Cloud console (the GUI), or the Google Cloud shell (a shell that has access to your temporary account / project) or even Terraform.

During all the labs, you’ll have a checkpoint button that will verify your progress – it usually works fine, but sometimes, it can take time to discover you did what was asked…

Labs are usually pretty easy, sometimes boring (repeating the same steps several times) you just need to follow through the steps. Except for the final lab of a series!!! This one does not guide you, and it can be challenging (but more interesting, as you need to apply your experience and previously acquired knowledge)

I completed 5 out of the 6 labs (I failed the last lab of a series 😞 and never re tried it… yet, cause I think I still can if I want with my remaining credits; talking about the credits, there are people actively looking for some, look at this Google Cloud community) and then my Google Cloud profile got updated with those badges, as well as another service named Credly that does link to your Google Account to display your badges too. It’s kind of confusing there are 2 different places to showcase your badges – just make sure you create your Credly account because this is where your Certification diploma will be uploaded.

After those labs were completed and those badges acquired, I had the opportunity to fill in a Google form to ask for a voucher to register for the exam.

Problem was it was already December 16th and my agenda only allowed me to choose December 24th for the exam date: I had 8 days (but truly only 3 cause I was working) to prepare for the exam 😱

Preparing for the exam

I decided to use external resources rather then the recorded sessions to fill the gap. I wanted to go through the maximum of mock exams.

Here are the external resources I used:

• Google Cloud Certified Professional Cloud Architect Study Guide: Second Edition, by Dan Sullivan – a bit outdated (published in 2022, by December 2024, many services were renamed – like Dataprep integrated into BigQuery – or even discontinued – like Stackdriver debugger) but I began realizing through some mock exams that I absolutely needed to learn, fast, a lot of services that were not covered in the labs!
• I quickly glanced through this other preparation book, Professional Cloud Architect Google Cloud Certification Guide – Second Edition, different style of questions, helpful too
• I spent a lot of time on Udemy Practice exams, I liked the questions even though the solutions are pretty barebone, with just a link to the doc…but after 3 or 4 mock exams, the questions were the same more or less.
• I also practiced on Whizlab, just the exam preparation tests – the solutions were better explained, but as all other resources, it was sometimes a bit deprecated

Among those 4 resources, I can not tell you which one better prepared me versus the other ones; I think with all of them I learned about new services or scenarios (a significant part of the exam preparation revolved around 4 case studies, where you need to know what’s the best GCP service , or combination of services – Pub/Sub + GCS + Dataflow + Big Query is always a winner 😊 – for a given business or technical requirement).

To gauge if you’re sufficiently prepared, make sure you don’t learn anything new when taking a mock exam – if so, you still have some learning and training to do.
Most mock exams gave more than 4 answers, or sometimes even allowed several correct answers: the real exam only ever had 4 answers, and only 1 to choose from.

Finally, there are a decent number of sites, blog posts on the internet about this certification; most of them a bit outdated, but they still helped:

• A good blog post with plenty of interesting links for the exam
• A list of other very good resources to learn about GCP services
• A list of services you should be familiar with before taking the exam

The exam

The exam is provided by CertMetrics and you can either take it online of in presence. I chose online because there was no availability in presence in Montréal in the next 2 weeks.

Online exam software

You should really follow their advice and use a personal computer, because they (CertMetrics / WebAssessor) will make you install a « OEMSafeBrowser » that is an evolved spyware to make sure you only access the exam and they can see and hear you during the whole exam (it’s no joke, during one question I repeated the answers aloud and they asked me to stop talking!).

You also need to be in a closed room where you’ll show them (using your webcam) you don’t have food nor any phones nor any papers around you!💂

Final note: I tried Firefox and Safari to start the exam (link to open the safe browser), supposedly they’re supported, but only Chrome worked – took me 15mn to figure this out, so I was late but hopefully the support delayed my exam. Still, that made me stress, so just use Chrome!

At the end of the exam, you’ll have a « conditional » result (after being asked to answer a satisfaction form!); I had the PASS result, so I was very much relieved 🥳 – few days later you’ll receive a Credly badge.

Types of Questions asked

20% of the 50 questions were about the 2 of the 4 case studies. But the questions were not the typical questions found in the online resources I used and they were not obvious either. Read carefully, every word matters.
The questions were only 4 answers quizzes, with only 1 answer to choose – that part was easier than most of the mock exam I prepared with (some of them had questions with 7 answers and 4 to choose!). The most similar exam was the mock exam from Google.

I found the questions to be quite difficult during the exam – about the renamed products, don’t worry about it, the questions use old and new naming.

Final words

Trying to get up to speed in 3 or 4 full days (not counting the labs) was extremely challenging, so try and prepare on a longer period than this, it’s not technically difficult, but there’s a lot of products to remember, many of them you’ll never have used during the labs, and have confusing and similar names (Cloud Dataflow is based on Apache Beam, but Cloud Composer is based on Apache AirFlow 🫠 – AppEngine Flexible is about running containers, the same as CloudRun – while AppEngine classic is a PaaS that supports a limited number of languages 😥 )

From my own experience (I am maintaining Java buildpacks , but used to be an application/cloud architect and devops), I have learnt a few interesting details on CloudStorage (similar but different from AWS S3) and Big Query and some privacy and AI services. I can’t say I’ve learnt much about application architecture, since it was pretty basic (« use managed services », « use PubSub and different zones to scale » etc. ) and the Big Data part was pretty limited to « use BigTable to insert, BigQuery to query analytics » 🤷 . Also, as an application developer, I can’t say I need to know that Cloud Interconnect is better than a (HA) VPN when peering with Google Cloud with 10Gbps needs or that Cloud Identity interfaces with Microsoft Active Directory… I don’t take this kind of company infrastructure decision.

Do certifications matter to you? to your employer? to your future employer? If so, go ahead (I was surprised to see sites such as Coursera, Udemy and Whizlab advertise that certifications will get you a better job or better salary – when I used to interview people, I didn’t really care about their certifications, I know it’s mostly theoretic knowledge that can’t beat experience)

I enjoyed most of the labs honestly, pretty interesting, but knowing about all Google Cloud services and their names of the moment…well I guess I now know what this certification is about!

Bonus: my notes

A list of all the notes I took and reviewed before the exam; could be helpful to you!