On November 28, Amazon announced it is entering the data warehousing field by announcing the Amazon Redshift service. Amazon Redshift is build on the ParAccel database platform, which can be accessed using SQL. Users can connect to the Amazon Redshift data warehouse using standard PostgreSQL JDBC or ODBC drivers. In addition, Amazon Redshift is already certified by Jaspersoft and MicroStrategy, two of the leading BI tools in the marketplace.
Amazon Redshift is a fully managed data warehouse. That means companies will not need to worry about time-consuming tasks such as capacity planning, data backup, monitoring, or patching. On the other hand, moving to Amazon does not mean that you can rely on Amazon for all your database needs. For example, if your query is running slow because an index is missing, Amazon Redshift will not be able to build that index automatically for you — a competent DBA is still needed.
Depending on the plan you purchase and how much you are willing to pay upfront, the price per Terabyte per year can be as low as $999. Similar to Amazon’s web hosting service, where the price is higher than shared hosting and low-end VPS but lower than dedicated server and high-end VPS, Redshift’s price point is higher than the budget system that small companies likely have. But when compared with what large organizations spend on data warehousing hardware, Amazon Redshift can be a very cost-effective solution. This is because many large organizations run their data warehouse system on top of databases such as Teradata, Netezza, and Greenplum, all of which offer excellent performance, though at pretty high prices.
As with all other cloud-based services, security is a concern with Amazon Redshift, as many companies will not be comfortable with moving its valuable data outside the confines of its firewall. Another challenge is data migration. Anyone who has ever done a large-scale data migration knows it is not an easy thing to accomplish.
My take is that, despite the buzz, initial adoption rate will be low. I can see some small- to medium-sized companies that are just starting their data warehousing efforts to use Amazon Redshift. For large enterprises, it will be a while before IT is comfortable with leaving data out in the cloud, and even then the initial projects are likely to be new data marts, as opposed to a full migration of the existing data warehouse.