AWS re:Invent 2015 recap: Freedom from constraints
Last week AWS held their annual re:Invent user conference in Las Vegas. This year was the largest re:Invent ever with over 19k attendees. Demand to attend was so high that the conference sold out of tickets months ahead of time.
An additional 30,000+ people tuned in to the watch the live stream of the keynote addresses where AWS executives Andy Jassy and Werner Vogels thematically set the stage of the coming year and announced new products and services. A lot of new products and services: AWS announced over 20 new product, services, and enhancements during the conference this year.
Let’s take a look at the keynote announcements from each day.
During his keynote address, Jassy positioned AWS as the fastest-growing Enterprise IT company in the world, with its core compute, storage and database services all sporting year-over-year growth rates of 100% or more. It’s remarkable that AWS is still achieving growth rates like this nine years after starting up its cloud.
Jassy’s keynote centered on how AWS now sees itself providing “freedom” to “makers” (the engineers, product designers, etc.) within a company that, in a pre-AWS world, were used to hearing an assortment of reasons as to why a new idea could not be pursued. In short, AWS is now in the business removing the word “no” from IT’s vocabulary.
Here are some of the themes I saw among the new products he announced.
Getting data into the cloud
The new offerings Amazon Kinesis Firehouse and AWS Import/Export Snowball aim to make it easier to get your data into AWS, whether by streaming it in one event at a time, or by literally FedEx-ing petabytes of data to AWS via a Snowball device. One thing that stood out to me about the new Snowball storage device was that you could use more than one in parallel—thus emptying out your data center in days instead of months or years.
Six letter word: Oracle
Jassy wasn’t very subtle when discussing the issues that people face using proprietary databases. He cited them as being one of the single the biggest pain points he hears from customers. It’s clear from news clips and the redacted logo he included in his presentation that when Jassy talks about his lack of love for proprietary databases, he’s really talking about Oracle.
This is a tightrope position to walk as AWS offers both Oracle and MySQL databases as part of its RDS offering. Perhaps that’s why his second group of announcements included AWS RDS for Maria DB (an open source RDBMS maintained by the founder of MySQL) and AWS Database Migration Service which makes it easy to move databases into the cloud and even switch databases types altogether. Again, a not-so-subtle hint.
Insight, not data
AWS users have traditionally had to rely on high-priced Business Intelligence tools to mine data housed in the cloud. Jassy’s introduction of Amazon QuickSight obviously changes that: AWS has stepped into the BI space. It will be interesting to see how partners such as Tableau, Tibco, and Domo work with the new service. However, more interesting to me is the integration with third party data sources such as Salesforce.com. Watch this space.
We all know that the only truly secure system is one that is unplugged and at the bottom of the ocean (or is it?). However, that isn’t stopping AWS from announcing some very cool security features including AWS Inspector, which takes a good look at your security posture; AWS Config Rules, which is a new rule-based system for enforcing compliance when it comes to resource usage; and AWS WAF, which is a much needed Web Application Firewall that will keeps your apps safe from the bad things that crawl around the dark portions of the Internet.
A few more…
AWS also announced a fully managed Elasticsearch Service for folks who don’t want to operate their own search engine infrastructure, as well as Amazon EC2 Dedicated Hosts for those of you who don’t want to share the underlying hardware your instances are running on.
Amazon CTO Werner Vogels gave the keynote on day two of the conference and, as usual, struck a more engineering-focused tone. “The cloud has removed all of the constraints,” Werner said, right before unleashing another ten new products and services. Here are some of the themes I spotted among his announcements.
Go big or go small
AWS offers nine instance families, each designed for different compute workloads. You’d think that with nine families that AWS had pretty much filled all the computing gaps. Not so. Werner announced the X1 instance family which is a giant instance with up to 2TB (yes you read that correctly) of memory designed for giant in-memory workloads such as SAP HANA. On the other extreme, he also announced the T2.nano: a tiny, burst-able CPU instance with 512MB of memory.
With the extremes covered AWS layered in another new pricing option for all instance: EC2 Spot Blocks. This spot pricing model is perfect for defined duration workloads such as batch jobs that can take up to six hours. This new spot offering allow you to save as much as 50% when compared to on demand usage. This is welcome addition considering that all EC2 Reserved Instances now require one or three year commitments.
The best server is no server at all
Despite all the new EC2 action at re:Invent, Werner advanced the theme of the best server being no server at all. On this front, he announced a series of enhancements to EC2 Container Service including a new Registry, CLI and AZ-aware scheduling. Perhaps even more exciting were the new features added to AWS Lambda which include the ability to run Python code, schedule execution of functions, create long-running processes (up to 5 minutes), as well as failure and retry logic.
With these announcements AWS is now fully running a three legged race when it comes to compute with three very viable compute models: instances, containers, and functions. Time will tell which of these models will be most successful for the various workloads and use cases currently supported by EC2.
Mobile is complicated
In a prior post I talked about how AWS is building an end-to-end mobile application platform. The newly announced AWS Mobile Hub is essentially the wrapper to that platform. It purports to simplify the entire process of building, testing and monitoring mobile applications built on AWS services. Think of it as a “wizard” that walks you through the dense juggles of AWS’s mobile capabilities and services, provisioning and scaling services for you. I think we are going to see a lot more mobile apps built on AWS.
That’s your refrigerator talking
Werner spent a large part of his keynote talking about how the Internet of Things (IoT) is the next big horizon for AWS. He painted the picture of a world where all manner of ordinary things (cars, engines, even refrigerators) are equipped with sensors that throw off huge amounts data. Werner and AWS clearly want see that data find its way into the AWS cloud, hence the announcement of AWS IoT—a framework that can be used to plug connected devices into AWS. The framework consists of device SDKs (embedded and Linux based), a message broker for handling the connections (it even speaks the ancient MQTT protocol), as well as a rules engine for acting on the messages and passing them to other AWS services for storage and processing. I suspect it won’t be long before my car is dumping data into S3.
It was another excellent year for AWS and everyone that attended re:Invent 2015. 19,000+ people definitely walked away from Las Vegas armed with new ways to build and manage in the AWS cloud. It will be exciting to watch how usage and adoption of all these new services will unfold throughout 2016. We’ll be sure to keep you updated!