Performance Tuning and Sizing Guidelines for Informatica® Big Data Management 10.2.2

Back Next

Case Study: Traditional Update Strategy versus Hive MERGE

The following case study compares the amount of time that an update strategy task requires to complete depending on the number of INSERT and UPDATE statements in the task and whether the task implements Hive MERGE.

Environment

Chipset	Intel® Xeon® CPU E5-2680 v3 @ 2.50GHz
Cores	12 cores
Memory	128 GB
Operating system	Red Hat Enterprise Linux 7.3
Hadoop cluster	14 nodes
Data volume	~ 75 GB

Performance Chart

The following performance chart shows a comparison of the time taken for an update strategy task to complete when the task contains a combination of INSERT and UPDATE statements:

Conclusions

Based on the case study, a 30% increase in performance is observed in the update strategy task when the task implements Hive MERGE.

When you perform incremental updates and the percentage of UPDATE statements is higher than the percentage of INSERT statements, consider using Hive MERGE.

Rename Saved Search

Table of Contents

Performance Tuning and Sizing Guidelines for Informatica® Big Data Management 10.2.2

Performance Tuning and Sizing Guidelines for Informatica® Big Data Management 10.2.2

Case Study: Traditional Update Strategy versus Hive MERGE

Case Study: Traditional Update Strategy versus Hive MERGE

Environment

Performance Chart

Conclusions