May 12, 2008, 9:37 AM — Database maker Vertica Systems
is moving its technology to Amazon's Elastic
Compute Cloud infrastructure (EC2), hoping to score customers who want a
hosted, pay-as-you-go model for data warehousing and BI (business intelligence),
the company announced Monday.
Vertica's database organizes data by columns, as opposed to rows. The company
and others that make columnar databases, such as Sybase and ParAccel, contend
the approach is faster and better for BI-related queries because only the desired
columns -- such as a customer's name or location -- can be read without having
to parse through an entire table, saving bandwidth.
The company also sells the database for on-premises use and in appliance form.
It sees a market for the on-demand offering due to a number of scenarios. For
example, a company might want to conduct a BI project that will only last a
fixed period of time, such as revising its pricing based on competitive and
market data, said Andy Ellicott, senior director of marketing.
Hedge funds, which test their trading algorithms against large sets of historical
stock market data, are another potential use case, because while such entities
manage vast amounts of money, they seek to maintain the lowest possible overhead,
which a cloud-based approach can provide, he said.
It is also hoping to sell to SaaS (software as a service) companies that focus
on analytics.
One such vendor, Sonian, has been beta-testing the on-demand version.
"Everything is working out really well," said Sonian's chief technology
officer, Greg Arnette. "We don't want to be in the plumbing business any
more, managing pipes and [storage area networks]. We don't have to become experts
on the Vertica database," he added.
The Dedham, Massachusetts, company started out as a hosted e-mail archiving
provider, but is working on a service that will provide analysis of a company's
stored content.
Sonian is still handling the task of loading data into Vertica, "but they're
providing the expertise on configuring the database, optimizing it, that kind
of stuff," he said.
Vertica's move is "very interesting" and positions it "as one
of the front-runners in the race to bring data warehousing into the cloud,"
said Forrester analyst James Kobielus in an e-mail message on Friday.
Amazon's architecture also sets up Vertica "both for the midmarket (which
generally doesn't use the most scalable premises-based [data warehousing] platforms)
and for the surge requirements of the most demanding, high-end enterprise DW,
predictive analytics and data mining jobs," Kobielus added.
Cloud-hosted database management systems (DMBS) hold "little value"
to "the end-user type IT and department-level
organizations," Gartner analyst Donald Feinberg said via e-mail.
However, there is potential in the model for small third-party software companies,
which could cheaply use the service to run proof-of-concept exercises, he noted.
"Another, although related to the software vendor, is when a BI app vendor
puts their app on top of Vertica and sells to their clients as a package,"
Feinberg wrote. "They could now sell just to the end-user (bypassing IT
and red tape) and install it fast with the correct resources for the end-users
on EC2, assist the end-user loading data and the users are off and running."
"So it appears that Vertica may be on to something for the short term,
until cloud DBMS (DaaS - DBMS as a Service) catches on more broadly (maybe or
at least two to five years out)," he added.
The on-demand offering is priced on a usage basis, beginning at US$2,000 per
month for managing 500 gigabytes of data. This compares to about $150,000 for
a typical on-premises installation, according to the company.
