Managing Synchronization Rules - Teradata Ecosystem Manager

Teradata® Ecosystem Manager User Guide

Teradata Ecosystem Manager
Release Number
December 2020
English (United States)
Last Update
Product Category
Analytical Ecosystem

Ecosystem Manager tracks table synchronization for tables with the same name and same database across multiple Teradata systems. Ecosystem Manager uses the sum of table activity received in events that specify the affected table, database, Teradata system, and metric type. The cumulative activity of each metric type is compared across Teradata systems for each table, and discrepancies determine the synchronization condition of a table.

The Ecosystem Manager has one pre-defined rule that states if there is a discrepancy across Teradata systems for any cumulative metric, for any table on any database, that the table condition changes to a warning level. You can add new rules, but they are exceptions to the pre-defined rule. It is recommended to keep rules to a minimum.

If you create rules with specific values for the database, table, and metrics, they take precedence over rules you create with wildcard (%) symbols. The following examples show cases for additional rules:

  • A table is updated on one system using incremental inserts, but the synchronization strategy is to copy the whole table to the second system using Data Mover. In this case, the INSERT metric for the table on the first system only reports the new inserted rows while the INSERT metric on the replicated system shows the total number of rows in the table. The metrics will never match for this table, so an exception rule is needed to remove this table from the sync. This allows Ecosystem Manager to skip the discrepancy and keeps the table in a normal condition.
  • If an entire table has metrics that will never match, the metric can be set to "%" in the synchronization rule. This means that all discrepancies for the table are ignored and table validation metrics are not compared, so use with care.

No alerts events are generated for table synchronization conditions since tables are expected to change frequently in most Ecosystem Manager environments. The table loading process creates temporary asynchronous conditions that are typically resolved and do not require action.