What Is a Unit of Work? - Teradata Ecosystem Manager

Teradata Ecosystem Manager User Guide

Product
Teradata Ecosystem Manager
Release Number
16.00
Published
December 2016
Language
English (United States)
Last Update
2018-03-29
dita:mapPath
qae1470238875438.ditamap
dita:ditavalPath
ft:empty
dita:id
B035-3201
lifecycle
previous
Product Category
Analytical Ecosystem

A unit of work is created during Ecosystem Manager instrumentation or is automatically generated by Ecosystem Manager later. A unit of work is used to track data applied to multiple Teradata Database systems, so you can determine if target databases are synchronized. The unit of work can be associated with one or more different jobs. The unit of work is a runtime parameter that is passed in the sendevent API for job events monitored by Ecosystem Manager.

A unit of work ID (UoW ID) is used to identify a specific workflow or unit and can contain the following:
  • File names
  • Date, timestamp, or time range
  • Tables, databases, or other database objects
  • All jobs
  • A data set
  • SQL extracts
  • Messages from a Message Bus (For example, JMS)
  • Any jobs or data that need to be defined
Examples of UoW IDs are:
  • Salesload090210 (job name with the date)
  • Productload092910063011 (job name with the date and timestamp)
Examples of unit of work processing are:
  • An extract job (inv_extract) is issued a unique UoW ID (inv_da7_030410). The job is run on an ETL server that is within the Grand Rapids data center. The same unit of work is used on a different ETL server in the San Antonio data center. Ecosystem Manager determines data synchronization by using the same UoW ID to track corresponding events in both data centers.
  • When a batch job occurs simultaneously on more than one data center, the UoW ID provides a way to compare results.
  • If the Grand Rapids data center has been down for 2 days, and the San Diego data center is up to date, the UoW ID is used in recovery and restoration processing.