AWS announce Redshift data warehousing service and price cuts
Amazon VP Andy Jassy takes to the stage at first AWS re: Invent conference to blast competitors private cloud offerings.
the market dominance of Amazon Web Services, it’s surprising to
learn that this is only the first year of their AWS re: Invent conference.
The three-day event in The Venetian Hotel in Las Vegas includes
keynotes, sessions from companies utilising AWS services and a
“Fireside Chat” with CEO Jeff Bezos.
At Wednesday’s keynote, Andy Jassy, Amazon senior VP in charge of
AWS, took to the stage to announce price cuts and a new data
warehousing service called Redshift, as well as taking a few
well-aimed snipes at competitors.
“When we launched AWS and cloud started to get some traction, we
used to hear old guard technology companies [saying]: there is
nothing new here. This is just virtualisation. There is nothing new
in the cloud,” said Jassy. “Customers saw through that.”
His criticisms could be seen as a response to Oracle’s Larry
Ellison, who famously described
cloud computing as “gibberish” and “insane”, before
an embarrassing about-face and the launch of Oracle
Jassy also blasted private clouds, accusing the “old guard” of
attempting to protect their traditionally high margins with sub-par
products. “You don’t get any elasticity, you don’t get that
innovation cycle, you still have to manage it and you can’t go
global in minutes,” he said.
He also provided a
few interesting stats regarding the sheer scale of AWS: S3 was
storing 1.3 trillion objects by the end of Q3 2012, while 3.7
Elastic MapReduce clusters have been launched since May 2010.
This scale is in part thanks to their aggressive pricing, which has
been reduced 26 times, said Jassy. This trend is set to continue on
December 1, when the price of S3 storage will be reduced by 24-27%
– perhaps in response to Google Cloud Storage’s own cuts
announced on Monday.
A shift in direction
Meanwhile, the company are continuing to innovate. The latest “big
thing”, unveiled during the opening keynote, is Amazon Redshift, a new data
warehousing service complementing its existing Relational Database
Service, as well as its more recent Hadoop service and NoSQL data
store, Elastic MapReduce and DynamoDB.
Data warehouses are traditionally used to process large quantities
of data for relationships, but are “too expensive and a pain in the
butt to manage” for large companies – including Amazon – according
to Jassy. In addition, he said, small companies don’t have the
resources to build their data warehouses, and existing data
warehouse-as-a-service offerings are inadequate.
Amazon CTO Werner Vogels goes into considerable technical detail
his blog, outlining some of the ways in which Redshift provides
speed increases of “10x – 150x”. These include storage of each
column sequentially, a massively parallel processing (MPP)
architecture and multiple connections to Amazon S3.
RedShift isn’t the only data warehousing service to launch at re:
Invent. A startup called BitYota emerged from stealth touting
a “Warehouse-as-a-Service for Big Data” which, in an ironic twist,
runs on AWS. Chief Executive Dev Patel
told the Wall Street Journal that, despite being taken by
surprise by the launch, he was confident that BitYota would be
“more flexible and easier to manage”.
Unfortunately for BitYota, it seems business intelligence companies
are already jumping onboard the Redshift bandwagon. Jaspersoft and
MicroStrategy have announced support, with additional business
intelligence tools “coming soon” to the platform.
The conference still has another day left to run, and Vogels will
be presenting his own keynote at 8.30am PST today. AWS may yet have
a couple more megatons left up their sleeve.