Skip to content

Latest commit

 

History

History

Folders and files

NameName
Last commit message
Last commit date

parent directory

..
 
 
 
 
 
 
 
 

README.md

JMH Benchmarks

Dependencies

How to Run

Generating Dataset

mvn compile exec:exec -Dexec.executable=java -Dexec.args="-classpath %classpath com.clickhouse.benchmark.data.DataSetGenerator \
-input sample_dataset.sql -name default -rows 10"

Running Benchmarks

With default settings :

mvn compile exec:exec

With custom measurement iterations:

mvn compile exec:exec -Dexec.executable=java -Dexec.args="-classpath %classpath com.clickhouse.benchmark.BenchmarkRunner -m 3"

Other options:

  • "-d" - dataset name or file path (like file://default.csv)
  • "-l" - dataset limits to test coma separated (ex.: -l 10000,10000)
  • "-m" - number of measurement iterations
  • "-t" - time in seconds per iteration
  • "-b" - benchmark mask coma separated. Ex.: -b writer,reader,i. Default : -b i,q
    • "all" - Run alpl benchmarks
    • "i" - InsertClient - insert operation benchmarks
    • "q" - QueryClient - query operation benchmarks
    • "ci" - ConcurrentInsertClient - concurrent version of insert benchmarks
    • "cq" - ConcurrentQueryClient - concurrent version of query benchmarks
    • "lz" - Compression - compression related benchmarks
    • "writer" - Serializer - serialization only logic benchmarks
    • "reader" - DeSerilalizer - deserialization only logic benchmarks
    • "mixed" - MixedWorkload
    • "jq" - JDBCQuery - query operations using JDBC
    • "ji" - JDBCInsert - insert operation using JDBC