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

Back Next

Deployment Criteria

The following criteria determine the Big Data Management deployment type:

Number of active users: Number of users working on the Model repository at design time, using the Analyst tool, or running Big Data Management jobs in the native or Hadoop run-time environment at any given point of time.
Number of concurrent pushdown mappings: Total number of mappings running on the Blaze, Spark, or Hive engines that are concurrently submitted to the Data Integration Service.
Number of objects in the Model repository: Total number of design-time and run-time objects in the Model repository. For example, data objects, mappings, workflows, and applications.
Number of deployed applications: Total number of applications deployed across all the Data Integration Services in the Informatica domain.
Number of objects per application: Total number of objects of all types that are deployed as part of a single application.
Total operational data volume: Total volume of data being processed in the Hadoop environment at any given point of time.
Total number of data nodes: Total number of data nodes in the Hadoop cluster.
yarn.nodemanager.resource.cpu-vcores: A property in the yarn-site.xml on the Hadoop cluster that specifies the number of virtual cores for containers.
yarn.nodemanager.resource.memory-mb: A property in the yarn-site.xml on the Hadoop cluster that specifies the maximum physical memory available for containers.