To thunderous applause, those of us not lucky enough to be there in person dragged ourselves out of bed at 3:30am to eagerly watch along this year’s re:Invent keynote from Adam Selipsky.
Like all other years, we were greeted by some classic music and we were all guessing what would be the theme of this year. Four words on the screen Vast, Unfathomable, Extreme and Possibilities.
So there it is, some random collection of 4 words. How will they connect to AWS’s assessment of the market, and all the goodies they’ve been waiting to release?
Lets start with Vast. In this case, Vast was an exploration of Space and Data.
The vast realm of data is our opening section where we hear from Adam talk about all the huge numbers from copious major customers – from 600 billion AI predictions at Expedia, to 80,000 requests per second at Samsung – really set the tone for how big data can be.
After talking through the range of Database and Analytic services we arrive at the first big announcement of the morning.
OpenSearch serverless
Preview: Amazon OpenSearch Serverless – Run Search and Analytics Workloads without Managing Clusters
We’ve got server less versions of all the other database engines, why not have an OpenSearch one. Serverless in this case looks to simply be reduced operational management of the clusters – allowing the AWS magic to automatically provision and scale out the underlying resources.
After the customer break we stay in data, but move into ETL – or as Adam put it “… thankless, unsustainable black hole”. Covering the different storage and analytic engines, we get offered a future with “zero-ETL” – maybe a pipedream, but one big pipe in there is now a little more free flowing.
Amazon Aurora zero-ETL integration with Amazon Redshift and Amazon redshift integration for Apache Spark
New – Amazon Redshift Integration with Apache Spark
With this integration, we now ease the pain of getting data from where it is, to where it needs to be – and reduce the time to insight. As commented by Colin Panisset – “Now we can leave the data where it is and analyse it with various integrated systems”.
Two big announcements in Data, but we are not done yet. Now that we have all of this data – how can we make it easier to share it, but also govern and control access to it.
Welcome Amazon DataZone
Pre-announcement, launch coming soon.
DataZone aims to provide fine grained controls to manage and govern access to data, populated from multiple data sources and overlay integrations into presentation and third party solutions.
This looks to be aimed more at the data consumer level, than LakeFormation and the data engineer level. We will be excited to take a look at this once its made available.
Next up in data is presentation – with two rapid fire announcements in QuickSight. In fact QuickSight is something I have recently had the opportunity to use, and Adam calls out that there have been over 80 enhancements over the past 12 months – with a lot of those in the past few weeks.
Quicksight Operational paginated reports
ML-powered forecasting with Q
Announcing Automated Data Preparation for Amazon QuickSight Q
While I’ve not used much of Q yet, the operational reporting enhancements are actually a big deal – if anyones had to try and produce a good looking report from a QuickSight analysis you’ll quickly understand why this is needed. Can’t wait to take it for a run and see how it goes.
This brings us to the end of data, and we move on into Unfathomable – here the analogy is the depths of the ocean, and how we need to adapt our senses to work in a darkness without vision.
The new world of Cloud has allowed us to build new ways to explore solutions, and Adam points out that Cloud makes it possible to be both Secure and agile.
This section is focused on their security suite of tools – everything from Identity & Access management, Network and Application protection through to Data protection and security tooling for detection and response.
Our first security announcement is the launch of Container runtime threat detection for GuardDuty – EKS has had GuardDuty support for a while now, this looks to be threat detection inside the container.
This looks to currently only be supported on EKS – hopefully its something that can be extended to support ECS in the near future.
Next up we look at the growing size of tooling in the security space, and again the volume of data that these tools create through events, metrics and logs. AWS have been working on an “Open Cybersecurity Schema Framework (OCSF)” with the industry to help defined some standards in interior on security events. To deal with this growing list of data sources, AWS announces:
Amazon Security Lake
Preview: Amazon Security Lake – A Purpose-Built Customer-Owned Data Lake Service
This one made me sit up – this is an actual problem I am dealing with at the moment – this is the kind of exciting re:Invent announcement that makes you login to your account at 4am and see if the service is available.
With a bit of url guessing, I was able to access the service and have already started on the journey exploring this. If it only gives me an “out of the box” view of VPC flow logs and ALB access logs, I will be very happy, but it promises so much more. Pulling in findings from Security Hub, Firewall Manager even the Health Dashboard – this looks to be a much broader operational lake than just Security.
And then like that we are out of Security and onto Extreme – here we journey to Antartica so that we can talk about ML? Yeah I kinda don’t get it either, but there must have been something there about Extreme levels of compute needed for ML training and inference. I was too busy playing with Security Lake.
I made it back to the Keynote to see the release of…
C7gn Instances for EC2 and Inf2 Instances for EC2
New Amazon EC2 Instance Types In the Works – C7gn, R7iz, and Hpc7g
And the hits kept coming with 2 new HPC workload instance types Hpc7g and Hpc6id – New – Amazon EC2 Hpc6id Instances Optimized for High Performance Computing
After the next customer break we return to an announcement that looks to be in the wrong section, but AWS make a play into spacial simulations with AWS SimSpace Weaver – New AWS SimSpace Weaver–Run Large-Scale Spatial Simulations in the Cloud
From the tin it looks to provide 3D simulation support into engines like Unreal and Unity.
I say the SimSpace announcement was in the wrong section, because our 4th and final destination is Possibilities – and this is a dive into the Imagination.
This is our “product” section of the keynote – where AWS have bundled their services together to give customers pre-built solutions to accelerate delivery. I personally really feel this is there area they need to invest in, and help technology buyers find pre-built solutions to meet innovation needs.
We kickoff with three new Amazon Connect features:
Agent scheduling | Agent performance management | Agent workspace to guide agents through scripts
Amazon Connect is a great example of what a Cloud Native mind can build using Cloud Native tools – I hope we see more of this type of innovation from the AWS folks.
We are on the home stretch now, thanks for sticking with me – and we’ll close out the summary as Adam did with 3 new business vertical solutions.
AWS Supply Chain
AWS Clean Rooms
No link available.
AWS Omics
These are 3 end-to-end cloud solutions to solve different business verticals. I expect the first two (Supply Chain and Clean Rooms) are internal Amazon.com solutions that are being white labeled for broader use. But I do see it as a positive trend in AWS putting their engineering where their mouth is and stumping up into the product game.
Only time will tell how well these launch, but great to see a broad set of feature announcements – some of which I need to get back into the console to take for a test drive.
See you at 3:30am tomorrow for the next Keynote.
Remember, to stay up to date throughout re:Invent 2022, sign up to our newsletter to gain exclusive access to highlights from the day, expert opinions, blog posts, video content and more.
