Amazon AWS held its first customer and partner conference last week in Las Vegas. So now that all the dust is starting to settle, what did we learn?
- It’s disruption time in the data warehouse and BI market. The biggest Amazon announcement at re:Invent was clearly Redshift. The new low-cost big data service is targeted squarely at legacy business intelligence and data warehousing worlds, with its clustered MPP architecture and ability to handle petabyte-scale data volumes. AWS claims that Redshift will be roughly 1/10th the cost of traditional data warehouses based on integrated software and hardware platforms from Oracle, Teradata, IBM and others. Interestingly Microstrategy and Jaspersoft have both certified the Redshift platform, reinforcing the fact that Amazon is seeking more to disrupt the infrastructure side of the market for enterprise data warehouse and BI solutions. As David Linthicum noted, it’s now Big Data’s turn to get “Amazoned.”
- The focus on enterprise continues to increase. Amazon also announced a new Data Pipeline data orchestration service that automates the aggregation across S3 and other data services, including EBS, Redshift and on-prem data repositories. The new service also enables scheduling of future data-related jobs, and back-up and data replication across locations. In a related announcement, Amazon also introduced new high memory and high storage EC2 instances, designed for intensive in-memory analytics workloads. In conjunction with the Redshift announcement its clear what type of customer Amazon is targeting in its crosshairs. (Hint – it’s not SMB.)
- Cloud Computing is still at the assembly-language phase. In a sign of increasing maturity, Amazon was open about where and how AWS needs to future mature to meet the needs of the enterprise. Reed Hastings, CEO of Netflix, one of Amazon’s flagship customers, cited the need for AWS to support the migration of live running instances, as is technically possible today with VMware and VMotion. Other AWS customers also noted the desire to actively migrate workloads around outages, and to automatically shut down inactive EC2 instances. Hastings compared the current state of cloud to his early days as a assembler programmer, noting that we’re still in the pre-compiler phase.
- The race to the bottom in commodity IaaS is accelerating. AWS announced price reductions on S3 by up to 24% depending on region. S3 now stores over 1 trillion objects, which demonstrates its scalability and broad usage. The latest price reduction comes right on the heels of a recent, previous reduction of more than 18% for first generation EC2 instance in US East and West regions. Not to be outdone, Google reduced pricing the week of re:Invent not once, but twice during the course of the week for a total of 35%. This downward acceleration of pricing is making life uncomfortable for two sets of people: 1) commodity IaaS service providers who are unable differentiate by moving up the stack or targeting market niches and 2) CIOs who may have built their private cloud business cases on faulty assumptions around the future public cloud pricing.
- Werner Vogels gets no love. The award for the quote of the conference had to go to Amazon CTO Werner Vogels, who shared in his keynote session that “I’ve tried hugging a lot of servers in my life and believe me, they don’t hug you back. They hate you.”
In addition to angry servers, I’m sure there are a lot of enterprise software and infrastructure vendors that also hate Amazon these days.