Analysis

Vertica ventures into the cloud

Philip Howard By: Philip Howard, Research Director - Data Management, Bloor Research
Published: 13th May 2008
Copyright Bloor Research © 2008
 Logo for Bloor Research

Vertica has just announced that it is offering its data warehousing technology via the Amazon Elastic Computing Cloud. This is quite a departure. While there are other data warehousing vendors offering hosted warehousing services, Vertica is the first to offer its facilities via cloud computing.

When I first heard about this I have to admit that I was dubious about how much take-up there might be but, having considered it, there are clearly a number of opportunities. The first is for companies offering analytics via a SaaS (software as a service) model: using a cloud-based approach means that they can extend their services without getting involved in building or extending their own data centre.

Secondly, there are small and medium sized companies who do not have the desire or the ability to build and manage a data warehouse of their own; and thirdly, there is a significant market within large enterprises that want a temporary data mart. Normally, you would load one of these up, query the hell out of it for a few weeks or months, and then close it down. But why go to the expense of having that kit hanging around, and the software licenses to go with it?

Finally, while it isn't a target market per se, a cloud offering gives Vertica a significant advantage when it comes to proofs of concept: implementation is more or less immediate and costs are minimal.

So, how does it all work? Well, actually it's trivial. There's no software to install or download; once you're signed up you start loading your data and once that's complete you can run your queries, analyses and whatever. You can opt for one, three or multiple processing nodes, as required. Provided you have three or more nodes then you get automated failover. Backup (which is highly compressed, to save on storage) comes as standard and, because Amazon is hosting the software you get Amazon level SLAs and support. Capacity starts at 500 Gb with a monthly fee of $2,000. Discounts apply once you get into the multi-terabyte range.

This is really quite appealing if you are in the right prospective user bracket. What is more, it is difficult to see some of Vertica's competitors offering the same sorts of capabilities as Vertica, as it requires an architecture predicated on the sort of grid technology used in cloud computing, which Vertica has but many others do not. Incidentally, there is no reason in principle why Vertica could not be implemented on other supplier's clouds.

I have to admit to not paying much attention to cloud computing prior to this so I am surprising myself by thinking that this looks like it has legs: I can easily envisage Vertica picking up significant numbers of implementations in a market that it effectively has to itself.

Reader Comments

Sorry, we are no longer accepting comments on this item. We suggest trying to contact the author directly.

16th May 2008: 'Jaspreet Singh' said:

I have been having my hands on Vertica for quite some time now & I can truly understand why Vertica came up with this offering on Amazon cloud.

Querying terabytes of data on Vertica is blazing fast as compared to any of the present DBMS's. Column store architecture gives them a natural advantage in terms of data compression since storing columns instead of rows increase the chances of data with same data type to be saved on disk. Hence data compression mechanisms are very effectively applicable to this data.

Vertica guys have build up an excellent product & are continuously refining the same. They just need a strong medium to showcase their product & what better way to facilitate easy proof of concept using Cloud computing apart from all the valid opportunities offered by cloud computing as mentioned by Philip.

Reply to Jaspreet Singh?

16th May 2008: 'Katie Elstree' said:

One thing eludes the reader when looking at Vertica's marketing collateral. Just how are users expected to get their data loaded into the cloud? By using the Internet? If so, do we really want a world where users are clogging up the Net's bandwidth by loading GB/TB of information. Surely this is the Achilles' Heel of this new solution. Telcos would hardly welcome this type of approach.

Reply to Katie Elstree?

13th June 2008: 'J Hammer' said:

The whole subject of analytics in the cloud debate does throw up some interesting questions: what about the integrity and security of companies’ data as they load it into the cloud environment? How many organisations are realistically going to let their data go into the Net? The gaping hole in the “analytics in the cloud” proposition is that regardless of the cloud computing providers’ size and position in the industry, HIPPA, Sarbanes-Oxley, and various other consumer data privacy laws will prevent many customers from moving their data from beyond their firewall. What fortune 100 company will to put their sensitive competitive information in a "cloud"?

Surely, what’s needed is a dedicated infrastructure, one that is set up for large, complex data analytics and that will take the pressure off the public domain.

It’s a debate that needs to happen before we all get brainwashed that this “analytics in the cloud” is the way forward for quick and easy data analytics.

The truth is that many organisations will still want to retain ownership and manageability of what can be very sensitive data. So how does the Vertica proposition get around that?

Reply to J Hammer?

The messages above were all contributed by IT-Director.com readers. Whilst we take care to remove any posts deemed inappropriate, we can take no responsibility for these comments. If you would like a comment removed please contact our editorial team.