View on GitHub

RadarGun

Benchmarking framework for data grids and distributed caches

Stage list

This document is related to RadarGun 2.1.x and contains list of all stages including their properties. XSD schema file containing information about stages can be found under $RADARGUN_HOME/schema/radargun-2.1.xsd.


add-result

Adds custom result to given test

analyze-test

Analyzes results of already executed test.

background-load-data-start

Allows to load data into a cache in the background, while other stages may take place. To force process termination, use BackgroundLoadDataStopStage.

background-load-data-stop

Stops data loading process started by BackgroundLoadDataStartStage.

background-statistics-start

Starts collection of statistics from background threads and cache size.

background-statistics-stop

Stop Statistics and return collected statistics to master.

background-stressors-check

Stage that checks the progress in background stressors and fails if something went wrong.

background-stressors-start

Starts background stressor threads.

background-stressors-stop

Stop BackgroundStressors.

basic-operations-test

Test using BasicOperations

bulk-operations-test

Executes operations from BulkOperations trait.

check-cache-data

Stage for checking presence or absence of data entered in other stages.

check-topology

Controls which topology events have (not) happened recently

clear-cache

Removes all data from the cache

cluster-validation

Verifies that the cluster is formed by injecting an entry into the cache and then reading it from other nodes.

command

Stage that allows you to execute generic command on the slave machine.

conditional-operations-test

Tests (atomic) conditional operations. Note that there is no put-if-absent-ratio- this operation is executed anytime the selected key does not have value.

define

Use for setting certain value

distributed-task

Stage which executes a MapReduce Task against all keys in the cache.

isolation-level-check

Stage for testing guaranties of isolation levels.

iterate

Iterates through all entries.

j-profiler

Stage for invoking operations on JProfiler. Remember to set up JVM args: "-agentpath:/path/to/libjprofilerti.so=offline,id=100,config=/path/to/configuration.xml"

jmx-cluster-validation

Validates formation of the cluster remotely via JMX.

jmx-cluster-validation-prepare

Collects configuration for JMXClusterValidationStage.

jmx-invocation

Allows to invoke JMX-exposed methods and attributes.

jvm-monitor-start

Starts collecting statistics locally on each slave node.

jvm-monitor-stop

Stop collecting statistics on each slave node and return collected statistics to the master node.

key-expiration-test

During execution, keys expire (entries are removed from the cache) and new keys are used.

load-data

Loads data into the cache using specified generators.

load-file

Loads the contents of a file into the cache.

map-reduce

Stage which executes a MapReduce Task against all keys in the cache.

monitor-start

Starts collecting statistics locally on each slave node.

monitor-stop

Stop collecting statistics on each slave node and return collected statistics to the master node.

parallel-start-stop

The stage start and stops some nodes concurrently (without waiting for each other).

periodic-heap-dump

Periodically generates heap dumps.

query

Stage which executes a Query using Infinispan-query API against all keys in the cache.

random-data

Generates random data to fill the cache.

register-listeners

Benchmark operations performance where cluster listenersTrait are enabled or disabled.

reindex

Runs Queryable.reindex()

repeat-begin

DO NOT USE DIRECTLY. This stage is added at the beginning of each repeat.

repeat-continue

DO NOT USE DIRECTLY. This stage is added at the end of each repeat.

repeat-end

DO NOT USE DIRECTLY. This stage is added at the end of each repeat.

scenario-cleanup

DO NOT USE DIRECTLY. This stage is automatically inserted after the last stage in each scenario. You can alter the properties in <cleanup/&gt element.

scenario-destroy

DO NOT USE DIRECTLY. This stage is automatically inserted after the last stage in each scenario.

scenario-init

DO NOT USE DIRECTLY. This stage is automatically inserted before the beginning of scenario.

service-start

Starts services on specified slaves

service-stop

Stops or kills (simulates node crash) one or more nodes.

set-log-level

Debugging stage: changes log priorities

set-partitions

Partitions the cluster into several parts that cannot communicate.

single-tx-check

Paired with SingleTXLoadStage. Checks that the previous stage had the expected result

single-tx-load

Performs single transaction with requests on key0 .. keyN in multiple threads on multiple slaves.

sleep

Sleeps specified number of milliseconds.

stack-trace-watchdog

Debug usage only. Periodically check for all thread stack traces and print them out.

tpcc-benchmark

Simulate the activities found in complex OLTP application environments.

tpcc-population

This stage shuld be run before the TpccBenchmarkStage.

wait-for-topology-event

Waits until some event occurs. Note that the initial rehash is not recorded in this manner, therefore waiting for that will result in timeout.

wait-for-topology-settle

Waits for a period without any change in membership/topology history.

write-skew-check

Stage checking the write skew detection in transactional caches.

xs-repl-check

Checks loaded data for validity. Useful for testing cross-site replication.