MemberDecember 18, 2020 at 10:37 pm
I’m currently working through the practice quizzes for the AWS DB Specialty cert. One of the questions is a multi-answer surrounding Performance Insights. Two of the options read “Performance Insights is configured at the cluster level” and “Performance Insights is configured at the instance level” and the explanations give you the information that it is enabled at the instance level and not the cluster level.
Well a few questions later the question is about finding reasons for CPU spikes and assessing queries. One of the answer choices is “Configure performance insights on the cluster” and I mentally eliminated that answer choice because you specifically say that performance insights can’t be enabled on a cluster and that’s the kind of phrasing that on other questions would mean that answer choice is wrong due to being inaccurate. Yet on this question it’s the correct choice….
I know this seems nitpicky or even perhaps that I’m merely salty over getting a question wrong, but if you have a test where some answers are considered “gotcha” or “trick” answers by seeming correct but being incorrect due to technicalities in the phrasing, then it seems inconsistent to play it fast and loose with phrasing on other questions. If you train people to eliminate trick answers by meticulously evaluating each detail and technicality, then having an answer that isn’t technically correct but is close enough in the general sense be the correct answer sends mixed signals. More importantly, you can’t specifically and explicitly say “Amazon RDS Performance Insights is configured on the cluster level.” is incorrect on question 10 because it’s configured at the instance level but “Configure Amazon RDS Performance Insights on the PostgreSQL cluster.” (these two answers are copy/pasted from the quiz) is correct on question 13 when normally the explanation for that answer choice would be something along the lines of “Performance insights is correct but you configure performance insights on a cluster not an instance”
Again, not trying to be nitpicky but when learning, consistency goes a long way.
MemberDecember 21, 2020 at 9:48 am
First, thank you for using our courses in Tutorials Dojo, and thank you for getting our AWS Database-Specialty course. Second, thank you for taking the time to share your feedback and your experience with the exams. We appreciate it. I hope that it has helped you get closer to achieving your goals.
I understand your main point and the arguments that follow. Consistency in the use of terms and jargon is very important. In response, we avoid constructing exams that intend to “trick” the exam taker for the sake of making it “difficult”. That brings little to no learning value. However, we do our best to make the exam experience as close as possible to a real AWS certification exam. What we want to emphasize is that the difficulty of certification exams heightens because of the use of language and the rules around it. It is very important to remember that this is also a language exam. My experience with exams for Oracle, Microsoft, and SAP since 2008 is very similar. We craft these questions with a genuine interest to challenge and evaluate the required areas of knowledge. Not just make it hard. Let’s dive deep now.
Amazon Performance Insights monitors database load and SQL activity. In PostgreSQL, regardless if it is in AWS or not, you monitor SQL activity by checking it inside the instance, not in the cluster.
One of the objectives of the first question (cluster level vs. per-instance level) is to test the taker, “Which step do you need to take to configure Performance Insights (and enable that monitoring) for a PostgreSQL?” In the AWS exam, they evaluate the taker’s capabilities and knowledge to implement the task via the AWS console, and sometimes even via CLI. Just like one of Amazon’s leadership principles, the focus on this question is “to dive deep” and be confident in the topic. Because of the use of “level”, we are checking at which AWS console page/AWS CLI command do you enable Performance insights. The answer is you enable it at the per-instance level.
The other question is focused on the concept of using Performance Insights to achieve the objective and solve the problem. It does not evaluate your knowledge to configure it, but your understanding of the various AWS offerings that can help monitor the database more effectively. Is the best solution to use CloudWatch, Lambda, native PostgreSQL options, or Amazon RDS Performance Insights? The choice “Configure Performance Insights on the cluster” does not focus on how you enable it. It simply means – enable it on all the instances of the cluster. If you enable it for a PostgreSQL instance of a cluster, that is incorrect because you do not capture all activities in the cluster. Your knowledge to accept that this choice implies it should be enabled for all instances of the cluster is critical to confirm that this is the correct answer. AWS exams evaluate the taker’s capabilities to find the best solution to solve the problem. It’s linked to another Amazon principle “Insist on the highest standards”. It is not connected with the previous question, and I would not recommend takers to associate past questions all the time. It helps when you forget some ideas, but it can get dangerous when overused.
What’s my final point? Take each question independently. Do not assume that one question is a precursor for another. Digest the fact that AWS preaches the Well-Architected Framework, and AWS, under Amazon, breathes the Amazon Leadership Principles. Their exams are testaments to how much they profess this and us at Tutorials Dojo attempt to construct these courses as consistently with these documents.
Hope this helps. If you have any other clarifications, happy to support you.
Joboy Pineda @TD
Log in to reply.