Recap from AWS re:Invent 2019 Day1

I'm on the ground in Las Vegas for re:Invent 2019. Sixty-five thousand people. 20+ minute waits for Uber / Lyft. Let's set the event logistics aside and hit the highnotes of Andy Jassy's 3-hour keynote this morning. Here's the least you need to know.

Overall impression: The keynote was long but actually well-managed for 3 hours. There were musical interludes by a cover band and a somewhat cohesive story around all the announcements. (Let's call most of the announcements "newish". You were already in the know if you read the AWS blog). I was invited to a keynote watch party by theCTOAdvisor - Keith Townsend. I highly recommend it. We had major dialogue in the room about what some of these announcements really mean for customers. We processed the implications of these items on K8S, serverless, software, databases. Folks in the room live, work and talk about this stuff all the time and we were overwhelmed. If you feel the same way you're in good company.

Here are some of the key themes I heard:

1. AWS starts from the premise that all workloads belong in the cloud - specifically in their cloud. They spent no time answering the "why" question or acknowledging that some workloads may need to stay where they are (onprem). He openly asked "Is that all you get for the money?" (referring to private cloud).
2. The end game of AWS is for you to run their stuff in their cloud. They're not just providing cloud services. They want you to consume their flavor of what you're looking for in their cloud. Andy did an interesting job of what I call "leading the witness" when he gave a maturity model on databases. Stage 1 is running a proprietary database like Oracle or SQL Server. Stage 2 is running an Open Source database like MySQL or PostgreSQL. Stage 3 is running AuroraDB - AWS's flavor of relational database. They tout that it resembles lots of what you like about Open Source with the performance of the proprietary stuff you need to leave behind. (Check out Nathan Avery's piece from 2018 on how Amazon is now a database company. His piece was almost prophetic.)
3. It felt like a startup died every 15 minutes in the keynote. Soo many of the announcements were cannibalizing companies who previously used to feed off of AWS.
4. AWS didn't reserve the punishment for startups. They threw major blows at many established vendors too. They came after mainframes, onprem cloud, Oracle, IBM, Microsoft, Intel, AMD and nVidia - if not be name then by implication.
5. Partner with AWS in a naive way to your detriment. If this were Star Trek - they'd be the Borg. If you partner with them (too closely) - you will be assimilated. They will study your IP and eventually beat you at your own game. We sold your offering yesterday. We're building a better version of it today. We'll focus on selling our version of that solution tomorrow.
6. There were also some very interesting things I'm curious to watch going forward. 1) Amazon Outposts are finally in GA - in native AWS but eventually in VMC on AWS flavor. (I'm not sure I like the idea of running VMC on AWS on Outpost onprem - especially if you're already running VMware in your datacenter. Feels a bit redundant). 2) Read up on the Aqua announcement with Redshift. The idea of moving compute to the data not the other way around is very compelling. It helps solve many of the challenges that come with data exploding all over the place. 3) I'm probably most skeptical AND intrigued about CodeGuru. The idea of being able to have a service to review code and quantify the cost of bad code is huge. It feels too good to be true. But starting down this path means we're getting closer to this being reality.

I give AWS and Andy Jassy credit. Being compelling for the better part of 3 hours is a herculean task. I predict that developers will be excited with all the new stuff they want to play with. Executives will be more overwhelmed with even more choices than before. Silicon providers like Intel, AMD and nVidia will be very concerned that AWS made a blatant push into their wheelhouse.

Quote of the day is courtesy of Keith Townsend from the keynote watch party:

"You can't ram what you do in the enterprise into AWS (as-is). Execs need to drive the change but don't have the depth of knowledge to drive it top-down"

Here's a Bonus take that almost all of us missed. Lots of stuff changed in Azure this week while all eyes were on re:Invent. Microsoft introduced some new configurations AND reversed some earlier price drops. You read that correctly. Some offerings went up as much as 60% from earlier price drops in late November.