Sign up for our newsletter and get the latest big data news and analysis.

5 Takeaways from re:Invent 2018. Because What Happens at re:Invent Doesn’t Stay at re:Invent

Last month AWS re:Invent attracted over 50,000 people to Las Vegas. With a “campus” spread across 7 hotels, there were over 2,100 breakout sessions that went deep into the AWS tech stack on topics like cloud architecture, continuous deployment, machine learning, and more. And from the moment you arrived at the Las Vegas airport, you knew AWS was going to be big—with bigger than life AWS signage plastered with hipster photography and captions like “the brains behind machine learning: that’s the next smart.” And there were well-executed conveniences too, like registration booths at the airport if you wanted to skip the line at the expo (or at the hotels.) Everyone who attended re:Invent probably groks what Jeff Barr of AWS meant on Twitch, when he said that re:Invent was “super, super fun—and incredibly exhausting.”

But what happened at re:Invent should not stay at re:Invent. My main takeaways:

“Builders are now a thing”

Fans of ‘Bob the Builder’ and the craftsmen who toiled to build the Duomo in Florence (for decades) would both agree: builders have always been a thing. And yet, this is the first tech conference I remember where the word “builders” was bandied about so much. Usually tech conferences focus their energy on “developers.”

During Andy Jassy’s keynote (Jassy is the CEO of Amazon Web Services), he defined “builders” as engineers, developers, engineering managers, operations managers, CIOs, chief digital officers, and CISOs. And then Jassy did his level best to make a connection to all of these different types of builders. Did his claims resonate, about what builders want, and what builders don’t want? Well, not with everyone. It’s just not possible to win over everyone. Still, when Jassy said that builders want freedom, superpowers, and to only pay for what they use—that resonated. Who doesn’t want superpowers?

Did AWS go full Red Wedding?

Amid the excitement about all the new technologies and innovations the AWS team announced, there was also a bit of angst. Bryan Cantrill of Joyent penned a dark tweet that made the rounds, in which he wondered when “reInvent [will go] full Red Wedding, locking the doors and announcing that every attendee’s product or service is now a forthcoming AWS offering. Or maybe that was this year?

The thing is, expanding your business by launching new services is a well-respected growth strategy. And as uncomfortable as it is when AWS competes with their own customers, this isn’t the first time it’s happened in the tech space nor will it be the last. Technology businesses of all shapes and sizes routinely engage in coopetition; heck, the neologism “coopetition” has been in the dictionary longer than most of us have been alive. Perhaps this is just the first time a cloud vendor has tried to expand into so many adjacent product spaces at this scale?

What’s a builder to do? “The broadest and deepest cloud platform” means choices, decisions, and trade-offs

The more interesting story for me was the approach the AWS team took to try to put a positive spin on the many choices they now present their customers, with over 140 services and 13 relational and non-relational databases. All these choices pose a problem for AWS prospects and customers alike: how do you choose? Amazon must know they have created a multi-headed monster of sorts, and that choice (usually a good thing) is causing their customers to have to work hard to understand all the services, features, and trade-offs. (a bad thing.) So the AWS team did the only smart thing they could do in the keynote: Jassy tried to transform AWS’s weakness (complexity of choice) into a strength, with the mantra that you need the right tool for the right job. An effective framing of the AWS situation, and one that I suspect many in the audience bought into.

Builders really do want the right database for the right workload

Jassy began the database section of his keynote by talking about the traditional “old guard database world” as a miserable, miserable place (think: Oracle.) And he was right, at least when it comes to the legacy Oracle world. And then Jassy went on to announce two new databases at re:Invent: the Quantum Ledger Database (QLDB) for blockchain and Timestream DB. And Jassy replayed the chorus of “the right tool for the right job.”

Developers do want the right database for their workload. But that doesn’t mean you need to deploy (and skill up on) different proprietary, purpose-built databases for each and every one of your workloads. Jassy is wrong in suggesting there is a big trend away from relational databases. And his own AWS statistics bely his prediction, since it’s an RDBMS database that has enjoyed the most growth on AWS.

The fact is, for most teams, developer time is the scarcest resource. And choosing a new and proprietary and purpose-built database when you don’t need one will waste your developer’s time, unnecessarily. And if these new purpose-built databases are proprietary or closed source—and if these new, purpose-built databases are cloud-only—well, that can cost you even more in terms of developer-time. With a cloud-only database, you cannot even run it on your laptop, or see the logs!

So this is where I disagree with Jassy. The relational database is still the right tool for the job for many modern workloads, no matter what new whiz-bang, proprietary, purpose-built databases have just launched. Is the relational database always the right tool? Of course not—but it is the right tool more often than the AWS keynote writers would like you to think. And yes, full disclosure, I have a bias: I agree with Marco Slot of Citus Data: the RDBMS is the future of distributed databases.

The unexpected takeaway from re:Invent—the merits of laptop privacy screens

Now that I’m back home and reflecting back on the whole week at the conference, I realize I have one more takeaway. If you travel on business, and you work on your flights, and you don’t have a privacy screen on your laptop: get one. During his keynote, Jassy told a story that he described as “a PR person’s worst nightmare.” It turns out one of his colleagues sat next to a senior person from a competitor on a flight out of Seattle. And during the flight, Jassy’s AWS colleague saw everything the competitor was working on. Hence: time for me to buy a privacy screen.

About the Author

Claire Giordano joined Citus Data to lead marketing and raise awareness about the worry-free Postgres database that is Citus. Before Citus, Claire served in leadership roles in engineering, product management, and marketing at Sun Microsystems, Quantum, and A9, an Amazon company.

 

Sign up for the free insideBIGDATA newsletter.

Leave a Comment

*

Resource Links: