Dave Brown is the VP of EC2 at AWS, which is too many acronyms. He declines to address the platypus in the room.
"So you run all of @awscloud EC2. I'm not saying that's a high stress job, but you *are* only 22 years old. How do you live with that kind of pressure?"
@primitivetype: "Maybe in hex. The trick is to have an amazing team."
Dave preaches the virtues of delegation. "@awscloud teaches all of us how to scale. Continuous scale isn't really optional here."
I point out that Dave had to do it the hard way and build this all from scratch. The point of cloud computing is that I don't have to use crappy things like a bunch of Raspberries Pi to build a site; I can start with @awscloud instances and keep going until I'm hyperscale.
"This is what I get for not properly vetting who they want me to speak with" thinks @primitivetype. He's far too kind to ever admit that, though.
I talk about the crappiness of early EC2. A single instance size (originally the m1.small), no persistent disk, etc. These days it's so much better on @awscloud. I talk about how modern instances are so much more reliable than the older instances.
Fun side note: when we stopped filming @primitivetype told me I'd stumbled over this upcoming release. OOPS! I think I found about four of those during the course of doing these interviews.
AFR or Average Failure Rate as AWS calculates it is whenever your EC2 instance does something atypical. It doesn't even need to be a visible-to-the-customer failure. And they're learning at scale that things are improving with time.
He further notes that nobody is deploying hardware at the same scale as @awscloud. I decline to comment on that because I don't want to find myself in the middle of that vendor war. HAVE FUN STORMING THE CASTLE!
We talk briefly about the spit-take awaiting a customer who just finished their migration from c4 to c5 instances in time to catch the c6i announcement.
Dave: "You comment on Twitter a fair bit about our deprecation policy."
Me: *oh crap he reads my Twitter feed*
Dave: "It's important not to break things for customers."
I assume there's one obstinate @awscloud customer on the legacy stuff. "Can we just move you to--" "NO!"
I take the time to praise the Savings Plan model over Reserved Instances. "The reason to migrate now is architectural and not dictated by economic considerations on a 3 year timescale."
I wish more things participated, like RDS. #awswishlist
I "helpfully" point out that 'Serverless' is a personal attack on @primitivetype. I speculate it was @ajassy unhappy with Dave's performance one year that led to the Lambda announcement in the first place.
Dave confirms that AWS services all live on top of EC2.
That's a big deal! "Is there some internal EC2-alike that runs your stuff?" "Absolutely not."
The fact that they use the same stuff internally that they sell us is incredibly important.
I reference this video. "I'm going to assume you just showed up at an Apple retail store one day and pulled a 'so, I'd like to buy some Mac Minis' and busted out the corporate credit card."
If so @primitivetype has enough Amex points to go to space now.
I reference that there's precedent to @awscloud buying Apple devices in bulk. Device Farm has a bunch of iOS and Android devices in data centers. It sounds like a prank but it's in fact real.
Since a lot of friends seem to be getting MacBooks Pro, I will dump some of the things I do to set mine up as a former Grumpy *nix Admin(tm).
A thread...
I've been using vcsh (github.com/RichiH/vcsh) for years to manage my dotfiles. I'm redoing my zsh config this break to remove over a decade of cruft. Early returns are looking promising.
# This function means `cdf` changes directory to that of your frontmost Finder window.
cdf() {
target=`osascript -e 'tell application "Finder" to get POSIX path of (target of front Finder window as text)'`
cd "$target"
}
Here's an unexpected thread from me; I never expected to write one quite like it...
A while back I had @ajyawn on the podcast, where he talked about @bytechek with me. It was a great episode, and I came away impressed by what AJ was doing. buff.ly/3pYnCGO
When I saw this tweet from him, I reached out to AJ with a "sounds like someone raised a funding round, and congratulations are in order." This isn't my first rodeo when it comes to reading the tea leaves.
I was excited enough about what @bytechek does (helps companies get to SOC2 compliance quickly, because I am a nerd as well as a former SOC2 control owner) and about @AjYawn as a person that I asked whether I could invest as well.
Because this is incredibly dense and technical, let me try to simplify it. I'm sure I will be condescendingly corrected if I get this wrong...
"We made a change internally that caused a bunch of internal things to become extremely chatty, like AWS employees defending the company if someone says something even slightly unflattering on Twitter."
So a question I posed in slack.lastweekinaws.com led to an unfortunate realization on my part:
@awscloud is too big, and has too many customers for the overall good of society.
"Well were things more reliable before @awscloud?" No! Good lord no! The difference is that I could have a bad day and take down a hospital. AWS has a bad day and takes down all the hospitals.
It's the simultaneous outage of everything that's the problem.
The worst part is that I don't even have the slightest clue how to fix it. You can plan and plan and plan around this. You can build out multi-region or multi-cloud until the cows come home.
And then one of your third parties did none of this and you're just as down.
Welcome to my first-ever livetweet of an @aselipsky#reinvent keynote as a part of my requinnvent.com coverage. It's 8:30, sarcastically loud, I haven't slept, and it's time to see what our @awscloud friends have worked on all year long.
A reminder: Snarking about companies is usually okay; snarking about people (presenters, etc) is not. Punch up, not down. Be kind.
The failure mode of "clever" is "asshole."
We begin with a jarring transition from "loud rock" to "easy listening combined with a Windows screensaver theme." Clearly the #reinvent graphics refresh was delayed due to... I dunno, not having a bad enough team name or something.
Staff: “Sorry, employees (orange lanyards) aren’t allowed in keynotes.”
Me: “It’s red. The lighting is super odd here.”
Once again I snuck past the “no Corey allowed in keynotes” rule!
And now I will livetweet the @awscloud partner keynote in this thread. #reInvent
@awscloud And now @dougyeum takes the stage and thanks us all for being here in person even though he won't be. "I'm moving to a new role within Amazon. Later, hosers."