QI Query Intelligence – Analysis & Tuning

As Teradata enterprise data warehouses become larger, more complex and increasingly mission-critical, the need to optimize queries has become crucial. A single poorly performing query can have severe consequences for the overall usability and operation of the warehouse, however freeing up operations and development staff to seek out high-impact problem queries can be a major challenge. QI implements Teradata performance analysis best practice and automates the process of finding these performance bottlenecks, and additionally provides the tools to remediate the problems in the shortest possible time.

A fully automated tool for searching DBQL on a regular automated and scheduled basis to identify queries that meet the criteria you’ve specified. QI runs autonomously, and notifies you when Queries-of-interest are identified. QI then analyses the queries and tells you, down to step-level, what’s good and bad about the query and what should be done to tune it.

• Creation of a profile of specific rule parameters that Teradata operations and Development staff want to automatically search DBQL for (i.e. Parallel Efficiency <60%, Query CPU >100,000 CPU Secs, PJI>7 etc.)
• Scheduling of QI to then run on a fully automated repeat-schedule basis against DBQL (i.e. once an hour, 9-5, Mon-Friday or whatever schedule is most useful)
• Automated identification of queries in DBQL that meet the profile rules you’ve created.
• Automated detailed analysis of Query Plans of any the queries that meet the created profile.
• Step-level information about identified queries of interest – the good, bad and ugly.
• Automated email alerts to nominated recipients for queries-of-interest identified by QI.
• Analysis results are also logged for further analysis and review, with additional dashboard capabilities provided, plus the saved files can be shared across the team.