|
|
Frequently Asked Questions
What is the Winter TopTen Program?
The Winter TopTen Program surveys users of large databases to understand the characteristics of the world's largest databases and the practices of the companies that build and operate them. Participants complete an online questionnaire about the size and complexity of their large databases, their hardware and software environment and configuration, the benefits derived, the problems encountered, and the expectations for growth their users have of these databases. (View the 2005 TopTen Survey.) In addition, respondents are required to validate key database scalability metrics by running database scripts to collect internal database statistics. Some of these metrics are published in TopTen lists on the WinterCorp web site.
The goals of the Winter TopTen program are to provide vendor-independent information that tracks the changing frontier of large database scalability and that describes some of the practices of the owners and operators of the world's largest databases. See the TopTen Program Overview for more information.
How does the TopTen Survey work?
Surveys are run periodically. The data collection period for the 2005 survey began in mid-February and ended on August 15th. The previous survey was conducted in 2003. During the data collection period - typically six months - respondents complete a web-based questionnaire and run a validation script that examines key internal database metrics. To qualify, the databases surveyed must meet a minimum size requirement. For the 2005 survey, the minimum size was 1 TB.
What’s the significance of “TopTen?”
“TopTen” refers to the ten winners recognized in each of the categories of the program. The first place winner in each of the 12 All Environments categories is recognized as the “Grand Prize” winner.
What does the Database Size metric measure?
Database size measures the actual storage in use for user tables, indices and aggregates. It excludes all temp tables log data, and database free space, as well as duplicated tables, RAID mirroring, and other redundancies to augment database availability. Database sizes are gathered from internal database statistics.
What does the Normalized Data Volume metric measure?
Normalized data volume estimates a database independent measure of the volume of user data which the database is managing, independent of database compression, indexing and other data transformations that affect the size of the data. To do this, the metric calculates the amount of data in the database by data type and translates it into a standardized format by data type. This information is gathered from internal database statistics and metadata, and sampling of actual data for variable length data types..
What does the Number of Rows metric measure?
Number of rows counts the total number of rows in user tables, omitting system tables and views for relational systems. For non-relational systems, the equivalent number of records is captured. This information is gathered from internal database statistics.
What does the Peak Workload metric measure?
Peak workload counts the number of SQL statements (or equivalent database operations in non-relational systems) executed in the peak hour of the measurement period. For queries, it measures the number of select statements executed, not the number of rows fetched. This information is gathered from internal database statistics.
How are these metrics validated?
Respondents run database scripts developed by WinterCorp in cooperation with the leading database vendors. These scripts gather information from internal database metadata and operating statistics. The scripts are designed to be quick, read-only, and non-invasive. Respondents submit the output of the scripts to WinterCorp. Respondents must complete both the online survey and the validation scripts to be included in the survey.
Why are no Teradata sites listed in the 2005 Survey results?
WinterCorp invites all large database sites worldwide to take part in the survey, and we work with their vendors to encourage participation. NCR Teradata decided not to participate in the 2005 TopTen program. In the absence of vendor recruitment, no Teradata users participated.
Are the identities of the program participants made public?
Winter Corp. publishes only the names, program metrics and database environment of the winning organizations. Companies may request to be listed as ‘Anonymous.’ See Past Program Winners for examples of participating organizations.
Are the results of the program made public?
Program findings are made public through several initiatives. WinterCorp. executes a media campaign that publicizes the winners and research findings. We’ll write a trade article, white paper and research report that features the TopTen winners and select program results.