Passed the GCE- PCA exam last week – experience & resource review.

GCE-Professional Cloud Architect certification experience.

TL:DR – the first few paragraphs are summary, the remaining pages are my breakdown of the exam guide after completing the exam.

Exam attempt success! It’s challenging, but at 50 questions its not long. I was able to complete the entire exam in 62 minutes. Marked ~20 questions for review, during the first pass. I spent the next 30 minutes reviewing the entire test, only made it about Q35; then just finished the marked questions. Ended up with 11 minutes left when I clicked the submit test button.

I enjoyed the way Google certification approaches their exams; the focus is on application of the knowledge specific to Google products. Though their can be overlap, the question describes the situation so you have to choose the best answer based on price, performance, reliability, scalability and support. There are at least 20 questions that required multiple answers, and 3 choices for many of them.

The use cases really make you feel like an architect solving customer needs using Google Cloud platform tools. The use cases deal with migrations/problem solving I have seen several times in my professional career.

Here is my review of exam guide and my interpretation of what you need to know.

Study Materials:

    Based on Brian post: Passed PCA at Google Cloud Next18

                Completed all 6 Coursera courses for the Google Cloud Architect Course.

                          This took ~40 hours of effort over a 3 week period.

                Watched all 70 Google Minute Videos per Brian’s Post.

I got pretty burned out completing the Coursera courses in 3 weeks. Beyond the Coursera course and the minute videos, I didn’t do any additional studying. The Coursera Program is Six courses, at least a week long, and 3 of the courses are multiple week. In my opinion, the course is lab heavy, being ~70% lab, 15% lecture, and 15% case study and quizzes. However, due to the labs, you understand and know the material.

Qwiklabs are great, and I found myself completing the labs in 30 minutes and at times using the additional 1.5 hours to launch any services I wanted. Kubernetes clusters, Marketplace services, databases, and Load balancers.  

Coursera had a special, so for every course, I was able to gain 30 free credits for Qwiklabs; I was able to get the full 180 credits :). I didn’t use any of these credits to prepare for the exam.

Cloud Architect: certificate exam guide review after passing the exam.

     Sample case studies:

          Review the 4 listed case studies before hand. Brainstorm different design scenarios and possibly re           architecting components to meet business objectives. I read them them through once just before the test; but due  to # of questions on these, I would have spent more time here.

By sections – things you should know:

1.0

You should understand cost and select the most economical way of deploying elastic workloads that scale with demand; ensure you understand how to scale down – much harder problem.

Cost for storage types and when they become beneficial. All google storage types offer the same performance, so the cost comes in redundancy – multi-region, or in access frequently – nearline (1m) and cold (1yr).

Understand the flow of data through load balancers, to web, to backend. Trade off between the types of storage used: persistent disk (local), multi-region, regional, nearline, coldline storage.

Know the tradeoffs between storage types.

Think about KPI’s, ROI, and which not only to collect the metrics, but how determine what you need to measure regarding the four use cases.

1.2

Understand how networks work (VPCs), load-balancers, instances and storage works across multi-region deployments for HA and failover design.

Understand App Engine, GKE, GCE, Cloud Functions and when to use each to meet an no-ops design.

Understand how to scale VM instances, kubernetes, managed instance groups, and also how to expand disk partitions in GCE instances.

1.3

Understand ingestion of Data into Google Cloud via: VPN, Dedicated interconnect, and how to overcome latency by using multiple VPN tunnels, or other strategies.

Understand which technologies you need for analytics and query: Should you use Cloud SQL, Spanner, DataFlow, DataStore, Dataproc, or BigQuery.

1.4

Implementing testing strategies for applications in test, dev, staging and production environments. Be able to determine how to test an application scale to meet SLA’s and implement HA designs to meet for SLA’s’

Understand how to connect on premises instances to GCP service, using service accounts.

1.5

Consider the growth of an application, ability to scale and planning to automate scaling and operations.

Ability to know which services other grow into.

2.1

VPN tunnels

Know firewall rules and methods of securing access.

2.2

Data retention and data life cycle management – commands via gsutil.

How to implement or migrate an ETL solution in real time.

2.3

Understand GCE, managed instances groups, kube clusters, autoscale groups, and insure you understand how these objects scale and the cli commands to do so.

Understand how to implement a shutdown script on a preemptible VM.

Understand what Jenkins is and it works.

Understand vpc, subnets, shared-vpc, sub-network.

3.1

Know IAM roles, but focus on service accounts

Understand – Org/Folder/project/Customer roles and the hierarchy.

Separation of duties/ Access / isolation between development and production using g-suite, projects, networks, or custom permissions between the different environments.

3.2

How to keep audit logs for 1 year or more. Stackdriver logging – export audit logs- Big query for analysis is an example or perhaps to cloud storage depending on additional requirements.

4.0

Continuous integration and continuous deployment.. Understand the difference.

Testing and validation of environments

Business continuity and disaster recovery..

4.2

Understand business requirements and also the objectives when making decision.

Create mental models of the data as it passed through network, compute, storage.

Chaos Monkey

5.0

Using App Engine with Cloud Endpoints to provide an API solution.

Managed the workflow of data in and out environments

Understand how to test solutions before the arrive in production. How to apply load test, scale testing, and debug testing.

5.2

Understand a docker file, json, yaml syntax. Know the requirements to run Java and Python and how to install both as dependencies for applications.

6.

Know Stackdriver

Distributed tracking setup

Logging export to cloud storage and big query.

Audit log tracking and storage

How to access logs for a specific instance in the Cloud console.

Know how to access network change logs

Good luck and I hope this helps the next person.  

Cheers,

Karlitos