I just sat for the exam and got certified in what is recognized as the 2nd hardest profession in Cloud stuff.
There will be a disclaimer and confession which will follow to state the intent of my post.
First, the prerequisites for taking the course were way beyond my paper and personal qualifications but the local provider wanted that nice govt money so they placed me in there. Fine. I don’t sleep anyway.
Second, I estimate less than 5% of my classmates who went and sat for the REAL exam that I did passed. The same fuckers who were giving me shit during a “capstone” project towards the end of the course, which was pathetic in every sense of the word.
I confess I contacted all the fuckers whom were giving me shit and asked them if they passed even though I KNEW THEY DIDN’T.
This is where the disclaimer comes in.
I am not smart. I just know how to spot questions because I was never looking at the subjects. I was looking at the business direction the cloud providor was trending towards. That already made it so I only needed to study 30% of the entire syllabus AND know what will NOT be tested. And I was, surprisingly, the only one out of a class of 40 year olds who was CONSCIOUSLY AWARE that the actual certification exam is not conducted by an INSTITUTE OF LEARNING but by a BIG FUCKING ASS MULTINATIONAL GLOBAL CORPORATION. So the questions wouldn’t be within the scope of an exam from the former even thought the latter had set the syllabus. So, I really only studied around 25% of the entire syllabus but did it WELL and FUCKING THOROUGH. I am really not doing that false modesty shit when I said I wasn’t smart as in academically inclined. I had 1/4 of shit less to study (which was pretty much a calculated risk) and 2 decades of business experience that’s all. And I am not in the least qualified, both in principle and in reality, if I were to go into the profession I’ve certified in. Irony. Love it lol.
Back to the confession. During the silly “capstone project”, I didn’t really want to give much of a fuck due to what I wrote above but some fuckheads got on my nerves when I casually said that the ones whom were taking lead were getting the “proposed solution” ass backwards. Was asked what I would do. I said I’d fucking go all out into Hadoop data migration and all the cloud native abstractions of HDFS and not the ones that were constantly being harped on again and again by the lecturer whom I do consider a mentor and have much respect for. He was doing his job. Beam is pretty damn amazing. The business objectives just didn’t warrant promoting it. But people hear Hadoop and it’s old school. Read about Twitter’s tech decisions and it’s like no one gives a shit about it anymore. Fuck. Read the STATS about actual usage by the majority of INDUSTRIES.
You see, cloud providors want enterprises to ditch on-premise stuff for the cloud. So they build stuff that are abstractions of what is used on-premise and manage the servers and stuff like resource allocation for them. So, focusing on learning their flagship services would seem logical. This is business. Business decisions are not anywhere near the decisions you make when you are looking to WORK FOR a business. You want to learn their flagship stuff so you will get employed for your knowledge of how to use them. That’s a smart move. And logical for graduates. But you fuckers are people in your 40s… and that’s not how business decisions work. They know their flagship services are fucking badass. What they want is CLIENT CONVERSION AND ADOPTION. Which means if you want to be in the profession, you need to know how to fucking get clients OFF what they are CURRENTLY USING. Not introduce something NEW AND BETTER. It does not work this way dudes. Not when your entire fucking business has invested 10s of millions and operated with the same fucking ecosphere and spent millions training employees to fucking maintain and expand their physical infrastructure.
You sell them the not-so-good shit that utilizes the fucking on-premise stuff they’ve built. How you can simply fucking do a push or pull of a data node from a HDFS cluster and spin the same jobs while saving a fuckload of money by not paying for processing power and storage at the same time.
No one but their tech leads know what parallel processing is, alright? No one making key decisions fucking cares that you can reuse the same scripts for batch and streaming pipelines. No one fucking gives a shit about what structured and unstructured data, let alone the cleaning, grouping, windowing, data warehousing, data lakes, DAGS, nodes, ML training methods, model pipelines, balls, etc.
I don’t normally gloat. But after giving me shit for 2 weeks and waiting 7 months for everyone to tell me I was right kinda gives me the fucking right to lol.