I'll never forget asking a guy who maintained 40yo mainframes for an insurance co why he didn't just migrate the apps.
His reply: the firm's actuarial code relies on specific bugs in the mainframe OS's floating-point handling. Run on any other system, it would produce chaos.
Absolutely fascinating insight I just heard from a hiring manager:
Adding junior engineers makes the senior engineers on his team *better*, because the constant "Why" questions force them to articulate and confront their own assumptions.
Hire juniors for a stronger team!
On Monday, I'll be starting a new adventure as head of content at
@googlecloud
.
Beyond excited for this chance to tell the story of the cloud at the biggest scale yet, in partnership with a visionary and brilliant group of colleagues.
Lately I’m seeing people defend the lack of junior cloud engineer positions by saying “cloud isn’t an entry-level job.” If that’s true, it’s quite an indictment of our industry, isn’t it?
I wish more CTOs understood that a "lift-and-shift" cloud migration is not a stepping stone to a better architecture; it's a delay tactic. And an expensive one.
I've been trying to figure out why this particular Discord message shook me so badly today. And I just realized why.
Scaring inexperienced young ppl into thinking they owe large sums of money - even if it's ultimately not true - is a life&death matter.
Hey AWS engineers! A friend's company is hiring a bunch of you. Fully remote, US-based, and they are open to folks who want to do part-time consulting as well as FTE. Very cool projects - serverless, IoT, etc. DM for an intro.
If I'm looking for a job 5 years from now, I'll be looking closely at how well $prospective_company retained staff between 2020 and 2022.
Anybody who's keeping their team happy and loyal right now, after the year we've all had, is doing something really right.
The term "skills gap" is a rhetorical trick that shifts the blame onto newbies. The actual problem is an "experience gap"; and that puts the blame right back on the industry for not providing onramps to experience for juniors.
I’m leaving Google Cloud at the end of this week.
Working at Google was something I’d dreamed of doing since I was 11. I met many great people in my 2.5 years there and got a taste of the complexity / impact of solving problems at hyperscale.
Grateful & excited for what's next!
My usual beef with reference architectures is that they launder the creators' assumptions into "best practices" without helping you put your own needs in context.
Turns out there's a cheat code to get past this problem, and the team that built "Emblem Giving" found it:
The 4 Horsemen of Serverless on AWS:
* Lambda
* IAM
* DynamoDB
* S3
Understand how to build and automate these 4 foundational services,
know their docs backward and forward,
and you can move the world.
I know
#hugops
causes some people to roll their eyes, but the Log4j thing is the reason it exists. A whole bunch of people are doing stupid amounts of stressful firefighting to protect us from a problem that they didn't cause. Thank you, ops teams. ❤️
I met a guy today who is transitioning from pest control to cybersecurity, and I think this may rival "commercial plumber -> DevOps engineer" as my all-time favorite tech origin story.
Some have pointed out the existence of AWS Educate Starter Accounts, which provide no-credit-card access to a limited subset of services.
Only via participating educational institutions, though.
Opening this up to all learners would be a huge step fwd.
The recent Roald Dahl thing is a reminder that if you care about your kids having access to great books, a physical home library is a smart investment. This is our kids’ collection, scrounged from Goodwill, library sales, etc - you can do this cheaply but the payoffs compound.
Unless something changes systemically at AWS, we should expect future Kinesis Incidents, and we should expect them to be progressively bigger in scope and harder to resolve.
My analysis of AWS's postmortem and the path forward:
My favorite question to ask the hiring manager when interviewing for a job:
"Can you tell me about a bad day your team recently had, and how you responded to it?"
Reveals so much about so. much. Every time.
AWS: “we don’t need developer advocates for IAM”
Me: *clarifies for yet another client the difference between an IAM user “account” and an AWS account*
Have to admit I got a little teary-eyed after
@werner
's
#reinvent
keynote ... here's a small, seasonal update to the S3 Ballad for the occasion. Big props to the AWS teams for everything you've built this year!
Today is my last day of full-time employment (at least for the foreseeable future).
I’ve been working toward this day for a long time, since I was an 18yo kid driving an IT service car covered in vehicle wrap.
Now I get the chance to do my own thing, on my terms.
Surreal.
5 things I've learned the hard way about multi-region / multi-cloud:
1. Building your own control plane, whether cross-region or cross-cloud, involves cost and complexity that you will have to deal with every day - not just when a managed service has downtime.
So, for the first 45 minutes of this
#reInvent
session I was nodding my head like "Yup, that's how I think about DynamoDB." Then Rick morphed into some kind of NoSQL wizard from outer space and my mind exploded. Absolute must watch:
This AWS war story is amazing, and perfectly illustrates the point I made in yesterday's newsletter. There is RADICAL labor savings (read $ savings) in the cloud if you commit to architecting properly.
From a salary + opportunities standpoint, learning to write has been worth >2x more to my career than learning to code.
In tech, code is table stakes. Effective writing and speaking set you apart.
I've finally achieved my dream of combining "Schoolhouse Rock" and cloud.
Enjoy "I'm Goin' To Production" (the Software Supply Chain Song) - ideally on the biggest speakers possible.
This is super interesting: AWS open-sourcing something close to the actual project layout, code, and ops config they use to run a serverless service in prod. Def. recommend reading the Github repo. Some surprises. May write up longer takeaway shortly...
After an incredible journey to acquisition, today is my last day at
@acloudguru
. Meeting Sam, Drew, and Pete at ServerlessConf Austin in 2017 changed my life. I will forever be grateful to have been a small part of one of the great teams in cloud.
Introducing the
#CloudResumeChallenge
. I'm volunteering my network to help you get your first job in the cloud. But I can only share a certain kind of resume. Thread ->
The reason people don’t hire juniors onto cloud teams is the same as it’s always been. Not because the job is too hard for juniors, but because the job of apprenticing and growing them is too hard for YOU.
I just don't buy this take that AI is going to decrease the absolute need for software engineers.
AI is going to generate SO MUCH CODE. We are entering a post-scarcity world for code. We have nowhere near enough engineers to edit, harden, maintain all the code AIs will generate.
People are out here pitching devtools to enterprises like "this will make your workflow easier and faster", and I'm like ... have you MET an enterprise?
Somebody called
@shanselman
the Bob Ross of programming and now that's all I can think of.
What he's doing with TikTok is brilliant. Humanity, earnestness and digital know-how - the great attributes of Gen Z - wrapped up in 60-second packages.
I vote we retire the term "non-technical" as a way to define non-software engineers.
The Excel wizard in your accounting dept or the CRO maven on the marketing team are as digitally-literate and automation-savvy as any programmer.
They just have less ego about it.
In the wake of The Kinesis Incident, I’d love to see AWS commit to a full audit of their internal service dependency tree and related assumptions. This write up, great as it is, does not give confidence in the blast radius of future cascading failures being less severe.
I'm hiring an experienced Product Manager to lead Google's investment in moving Alphabet & many Google products to Google Cloud. It is a fascinating role that will shape the future of Cloud.
Role says Canada but anywhere in North America is good.
Next-gen Google Cloud Functions just dropped. Most interesting to me: this version is built on top of Google Cloud Run. Stay tuned for a behind-the-scenes deep dive on this architecture soon.
People who are fluent on more than one cloud provider - how has this skillset affected your career?
Does it mostly just expand your job opportunities (ie, you can get hired by a GCP shop or an AWS shop), or do you find yourself working with multiple clouds at your current job?
Hey, I am HIRING for multiple roles on my team at
@acloudguru
. If you have a background in tech and a knack for journalism, I want to talk to you. DMs open.
Another blog I need to write up soon: "Why every engineer should do a stint in consulting."
• Enforces clear communication, esp. in docs
• Builds empathy for user needs
• Reveals the business impact of code
• Improves "mean time to competence" on new skills
@JoeEmison
To be clear, I don’t think he was giving me an excuse for why the code couldn’t be refactored - he was explaining why it couldn’t be lifted&shifted as-is.
Getting to play a small role in stories like this is what keeps me involved in tech.
Here's a breakdown of how a plumber with zero cloud experience got hired off of the most compelling entry-level resume I've ever seen:
"Serverless computing will become the default computing paradigm of the Cloud Era." Here is a rare academic paper that everybody in the industry should read:
The fundamental error all of these GenAI use cases make is in assuming that people will want to read something that other people couldn't be bothered to write.
Still thinking about
@mipsytipsy
’s Future of Ops talk. What really stood out: great engineers are a product of great teams, not vice versa.
Within 3-6 months, even the brightest new hire usually rises/falls to the level of the surrounding system.
This hurts, but rings true.