The last time I was fortunate enough to attend AWS’s global conference, re:Invent, was three years ago in 2014. Then there were 14,000 delegates and the conference spanned just two Las Vegas hotels. Lambda was announced during Werner Vogels’ keynote and it seemed that the most in-demand sessions had “Docker” in the title.
In just three years the conference has tripled in size with 43,000 delegates attending this year spread across a campus of six Las Vegas hotels. Although not one of the biggest conferences held in Vegas, it’s obviously a significant logistical challenge. After some hiccups on the first day with the inter-venue shuttles and a venue running out of food, everything seemed to settle down and run smoothly from the start of the second day. Whether the improvement was due to human learnings of the hivemind or training of some Machine Learning algorithms is up for debate but almost certainly it was a combination of the two. No, actually, the transport still is not good and Uber is key to success.
An Abundance of Choice
One of the consequences of attending an event of this size is being utterly spoiled for choice and overwhelmed with inspiration. Getting to and into the sessions you planned for can be hit and miss. All you can do is keep trying and hope that serendipity is kind. Then there’s also the dilemma on whether to go to the “immediately applicable” or “work related” sessions, or the “That’s so cool!” sessions on newer services and features. When the answer is “all of the above please” the degree of scheduling difficulty is raised again
The trick though seems to be trying to avoid sessions that have a high recap-to-new-information ratio, but this is not always easy when it comes to more established services or topics. I wouldn’t expect a 300 level Elastic Block Storage (EBS) Deep Dive to run through the volume types without adding much, for example.
On the plus side though, all of the breakout sessions are recorded and posted on YouTube so if you miss one that you later find out was “unmissable”, you can catch up. There’s possibly more value to be found in the interactive displays, events, workshops, hackathons and jams, and of course the social programme, anyway. It’s been great to catch up with old friends from afar and hang out with colleagues from other Melbourne-based DevOps shops, and the conversations always take interesting turns once the drinks begin to flow.
The Rise of Lambda
After the first two days of breakout sessions I attended it became obvious that just three years after its announcement, Lambda has become ubiquitous within the AWS ecosystem. I don’t think I attended a single session where Lambda wasn’t either central to the solution or else the cream on top. Even that EBS deep dive session had a cool automation using Lamda. Of course Lambda is a key part of “Event Driven Serverless” Architectures and as Andy Jassy just reminded us in his keynote, AWS have no fewer than 16 event sources and are clearly charging in the Serverless direction, notably with Serverless Aurora also announced today.
More about the Keynote
There are better places than this post for an exhaustive list of the announcements made during Andy Jassy’s keynote this morning including the @AWSreInvent twitter feed and Nic’s great work in #reinvent2017. We filed in past a very hyperactive and sweaty DJ who must have done the equivalent of a whole hour’s workout before the keynote began.
Andy came out and made some really significant announcements in the database, container orchestration and machine learning fields. There was realtime video stream analysis with Rekognition for video and Kinesis video streams, language Transcribe, Translate and Comprehend, Sagemaker, freeRTOS, Fargate and it goes on. SELECT for S3 and Glacier Data Lakes… Kubernetes as a service (EKS) was probably the most anticipated (aka worst kept secret) of the conference and there were jibes at Oracle from the Stage and obvious gauntlets tossed towards Google and Microsoft.
The keynote was certainly a very slick production with a “house band” playing covers with pertinent, to Andy’s presentation, lyrics. We had Lauryn Hill (Everything is everything), George Michael (Freedom) and Foo Fighters (Congregation). Lots of attention and nods to “Freedom” (this was where Oracle got slammed) and “The Makers” with themes of blind faith and false hope juxtaposed and parlayed into Analytics.
On Tuesday I’d attended a session on IoT, Rekognition, Lambda, SNS etc. for analysing IP Cam feeds and alerting with meaningful data. Rekognition can provide all sorts of data when fed images from, for example, an S3 bucket by S3 events. Object tagging, facial recognition, analysis and comparison, sentiment analysis and even moderation tagging are all available today and no doubt more data types will come.
Not long ago I setup an IP Cam at an Airbnb property and connected it to Zoneminder on a Linux host for capture and motion detection. So version 2 of my IP Cam setup beckoned as a new project after attending the IoT/Rekogniton session. Still looking forward to that project but now, as of today, we have DeepLens hardware for helping to get more “Everyday Developers” into ML.
The Lure of DeepLens
I’ve already attended a session in the ML track so will receive my free DeepLens in the post “early in the new year” but right now I am trying every way I can to get into a DeepLens workshop and grab myself one while I’m here! They’re the hottest ticket right now and I may just have to “catch up” on some of those DevSecOps and Containerisation sessions in favour of a DeepLens workshop. Best laid plans and all that…and I have to get to the IoT makers display too. So little time… (Update: There was such demand for the workshops offering these that they brought in extra security!)
….and that’s a wrap (for today)
So as reInvent 2017 rolls on for the remainder of the week we have Werner’s keynote, more of all session types, and probably some hot newly announced sessions to distract us all from what we originally planned. Feels like it will continue to be very busy and interesting. This conference isn’t over yet, but I already know I will sleep well on the ‘plane home!