Managing the Data Lifecycle by Jeff Maresh

There are a number of ominous signs to warn you that your data warehouse lifecycles are not being managed well. One sign is that you get to be real good friends with your data storage vendor. You have lunch at a nice restaurant a couple of times each month to discuss storage needs and they take you on golf outings every quarter. You experience this generosity because your quarterly orders for their products are measured in multi-terabyte chunks. These lunches are nice respites for you and your staff who work nearly 24x7 at a frenetic pace just to keep the warehouse running. That backup that is now taking nearly four days to complete is another warning sign, particularly since just five months ago you installed a new high-end tape storage subsystem and paid the vendor good money to optimize all of the backup software. Nine months ago, you increased the number of CPUs on the system by 50% just to keep end-users happy with query response time. Now you find yourself negotiating for a brand new replacement box with a lot more horsepower. How much worse can it get?

 

Successful data warehouses accumulate large volumes of data over time. The majority of the data are the quantifiable measures that occur everyday in the business that accumulate in fact tables. Shortly after a business event occurs, the measures find their way into one or more fact tables and users have a lot of interest in them. Over time as more facts are added, the older data become less useful for answering business questions, so considerably fewer queries are run against them. Finally after some period of time, the data are no longer of any use. This is the data lifecycle and a single data warehouse will have many different ones associated with the various fact tables. This paper discusses how to effectively and efficiently manage data lifecycles.

 

When asked to "Select a payment method" for your free download, choose "Print and Call". Your download code will be e-mailed to you after we authorize it. International customers should enter TruTek's address as the shipping address and enter their real shipping address in the Special Instructions field on the payment screen.

  • Author:: Jeff Maresh
Managing the Data Lifecycle by Jeff Maresh
  • Item #: Paper0020
  * Marked fields are required.
Qty*
Price $0.00
Availability In-Stock