The creatively named AWS re:Invent 2015 is now in our past. It’s clear AWS was charging at the enterprise mainstream with a big focus on database migration and adding additional business analytics capability. IoT was also a big push. But perhaps the energy at the conference is a trailing indicator? What if the post-AWS world is already in front of us with container-centric platforms that can run on any cloud, and AWS is relegated to being the king of IaaS?
- 3 Takeaways from Amazon’s re:Invent cloud conference by Brandon Butler at NetworkWorld
- 10 highlights from AWS re:Invent 2015 that cloud experts must know by Janakiram MSV at TechRepublic
- Real world lessons from AWS re:Invent 2015 by Bernard Golden at CIO
- Amazon’s Anti Gravity Play: Notes on AWS re:Invent 2015 by James Governor at RedMonk
- AWS re:Invent and the risk of filling in the white space by Ben Kepes at Computerworld
Along with the bigger enterprise-ready and growth stories, I see two big product themes:
All Your Database Are Belong To Us
Having pwned all your servers, it seems like AWS wants all your databases and your data analysis. This is seen as a threat to everybody from Oracle and Microsoft to Cognos and Tableau to Google Cloud Platform.
The Import/Export Snowball got a lot of press, and it is cool. This 10TB tamper-proof appliance is the proverbial station wagon full of tapes. For those who are don’t think the Snowball is stupid, I’m reminded of the original reactions to to iPod 14 years ago – don’t discount how much whole-solution and ease-of-use count for product adoption today.
I am also impressed that 1000 customers signed up in the first week alone for the AWS Database Migration Service.
Internet of AWS-Connected Things
AWS and IoT seem like they’d go together like pancakes and peanut butter. Lots of good descriptions of what was announced on the news sites — secure message queues, a device state db, and some hardware alliances. It’s early days yet with IoT, and I’m withholding judgment until I find some good IoT practitioners to weigh in for us.
Are we at Peak AWS?
I know Gartner says AWS has as much revenue as its next 14 competitors, combined. So that’s game over, right? As we know, the thing that kills you doesn’t look like you (e.g., nobody had a better hypervisor than VMware, but AWS is not a hypervisor; likewise, nobody’s going to be a better AWS than AWS). Keep your eye out for the following – if AWS was going to start to be eclipsed by the next generation, some of these things could be in play:
- Yes, AWS innovates, but are they better at launching new services than improving the old ones? (See this 4-year-old G+ post from Steve Yegge, where he praises AWS’s platform stance but makes this very criticism.)
- AWS was able to establish an operational shift (i.e., moving your servers up to the cloud), but can they establish a developer shift? Joe Beda on Operational vs Developer Lock-In. I haven’t seen anybody writing about this, but especially as AWS moves into everything from business intelligence to virtual desktops, they can’t be best-of-breed at everything. Especially for enterprise apps, being fully-featured and mature is still heavily valued.
- Related to the above, is a post-AWS compute stack emerging with containers? Sure, Amazon has a set of platform services, but there are alternative platforms like Cloud Foundry or OpenShift (or components like Mesos or Kubernetes or Swarm) that have plenty of ecosystem weight. With the new app stack, is AWS just one choice among many? Joe Beda again on what this new stack might look like: Anatomy of a Modern Production Stack.
- And finally, I keep hearing whisper stories about people moving back from AWS to private data centers, but few stories on the record. Real trend or wishful thinking from the old guard?