Classification Settings - Teradata Viewpoint - Teradata Workload Management

Teradata® Viewpoint User Guide

Product
Teradata Viewpoint
Teradata Workload Management
Release Number
16.50
16.20
Published
July 2021
Language
English (United States)
Last Update
2021-07-27
dita:mapPath
yqb1520516846011.ditamap
dita:ditavalPath
yqb1520516846011.ditaval
dita:id
B035-2206
lifecycle
previous
Product Category
Analytical Ecosystem

Workload Designer provides a common classification process for workloads, filters, throttles, query sessions, and utility sessions on the Classification tab. Classification determines which queries use which rules. Teradata Database detects classification criteria before executing queries. The goal in creating a useful classification scheme is to meet business goals and fine-tune control of Teradata Database. The following classification criteria types are available:

Request source
Where the query comes from, such as username, account name, account string, profile, application, client IP address, or client ID.
Target
What the query is acting on, such as database, table, view, macro, stored procedure, function, or method.
Query characteristics
What the query is composed of, such as statement type, AMP limits, any step time exceeds, step row count, final row count, estimated processing time, minimum step time, join type, full table scan, memory usage, or incremental planning.
Query band
What metadata is attached to the query, such as user location or application version.
Utility
Which utility submitted the query, such as FastLoad, FastExport, or Backup Utilities.

You can modify the classification settings in response to data monitoring, regular historical analysis, or changes. For example, classification groups may need to be created, or existing groups modified, if an application is added, two Teradata systems are consolidated, or service-level goals are missed.

A good approach to using classification is to first use request source to determine where the query is coming from. Often the account string is selected, but other options include username, account name, or client IP address. If you need a more detailed level of classification, establish where the query data is located, such as a database, table, or view. To narrow classification further, select query characteristics, query bands, or utilities. For utilities, use the check boxes to select the specific utilities you want to include.

For example, you could create a filter and add the request source classification to reject all queries from the Finance department when the Red state is in effect. To further refine the filter, add the query characteristic classification to filter out all requests from the Finance department that are estimated to run longer than 10 seconds. You include classification items in a filter to reject those items. The query characteristic setting is added to the request source setting already in place.

Generally AND logic is used when you add more than one criterion, so that the definition becomes more specific as more criteria are added. However, if you add both the Username and Profile request sources, the relationship between the two can be specified as AND or OR.