On December 9, 2009 CloverETL Cluster Edition was launched at PriceWaterhouseCoopers premises. CloverETL Cluster intelligently partitions data and distributes them evenly across multiple nodes in a cluster for execution in parallel. CloverETL Cluster’s ability to load balance large data transformations increases throughput, fault tolerance and flexibility.
CloverETL Cluster can be deployed on premise in a customer’s own data center or in a variety of cloud configurations, such as Amazon EC2, which can drive costs even lower. During the launch in Prague, CloverETL Cluster was demonstrated running on four Amazon EC2 servers.
The following table shows the time CloverETL requires to execute a moderately complex transformation of six million records (725 MB) in a variety of local and cloud configurations:
- CloverETL Desktop Designer on a MacBook: 150 seconds
- CloverETL Cluster Load Balanced Across Two EC2 servers: 60 seconds
- CloverETL Cluster Load Balanced Across Three EC2 servers: 43 seconds
- CloverETL Cluster Load Balanced Across Four EC2 servers: 31 seconds