[CASSANDRA-11542] Create a benchmark to compare HDFS and Cassandra bulk read times

• Type: Sub-task
• Status: Resolved
• Priority: Normal
• Resolution: Fixed
• Component/s: Legacy/Testing
• Assignee: Stefania Alborghetti
• Reporter: Stefania Alborghetti

Authors

Stefania Alborghetti

Reviewers

Russell Spitzer

• Created: 11/Apr/16 03:40
• Updated: 16/Apr/19 09:30
• Resolved: 25/May/16 08:29

Description

I propose creating a benchmark for comparing Cassandra and HDFS bulk reading performance. Simple Spark queries will be performed on data stored in HDFS or Cassandra, and the entire duration will be measured. An example query would be the max or min of a column or a count(*).

This benchmark should allow determining the impact of:

• partition size
• number of clustering columns
• number of value columns (cells)

Issue Links

• relates to: CASSANDRA-11521 Implement streaming for bulk read requests
• links to: Benchmark code
• links to: cassandra modifications (poc)
• links to: java-driver modifications
• links to: spark-connector modifications