The benchmark set contains 9 workloads. They fall into two categories. The first category is "simple resource benchmark", the goal is to test how storm performs under pressure of certain resource. The second category is to measure how storm performs in real-life typical use cases.
-
Simple resource benchmarks:
- wordcount, CPU sensitive
- sol, network sensitive
- rollingsort, memory sensitive
-
Typical use-case benchmark:
- rollingcount
- trident
- uniquevisitor
- pageview
- grep
- dataclean
- drpc
- Build.
First, build storm-benchmark.
git clone https://github.com/manuzhang/storm-benchmark.git
mvn package- Config. modify
./benchmark/conf/config.sh.
# Where is the storm binary
BIN=/usr/lib/storm/bin/storm
# Absolute path to the stom-benchmark-with-dependencies.jar
JAR=/root/storm-benchmark-0.1.0-jar-with-dependencies.jar
# Please don't modify this
MAIN_CLASS=storm.benchmark.tools.Runner
# We will pull the metrics from nimbus periodically. This defines the interval.
METRICS_POLL_INTERVAL=60000 # 60 secs
# How long will we run for each benchmark.
METRICS_TOTAL_TIME=300000 # 5 mins
# Where we store the metrics reports. The metrics contains the performance and throughput information.
METRICS_PATH=/root/benchmark/reports
METRICS_CONF=metrics.time=$METRICS_TOTAL_TIME,metrics.poll=$METRICS_POLL_INTERVAL,metrics.path=$METRICS_PATH
# The default workers we use for the benchmark.
WORKERS=4
# The default ack tasks we use for the benchmark.
ACKERS=$WORKERS
# The default max.spout.pending(it will override the default storm config) we use for the benchmarks.
PENDING=200
### the kafka configuration
# Kafka broker list [node1:port1, node2:port2, ...]
BROKER_LIST=intelidh-04:9092
# Zookeeper server list [node1:port1, node2:port2, ...]
ZOOKEEPER_SERVERS=intelidh-04:2181
# the root path you create for Kafka in Zookeeper
KAFKA_ROOT_PATH=/kafka/kafka-cluster-0 - Choose. Pick the benchmarks to run in
benchmark/conf/benchmark.lst, for example, to run wordcount only
wordcount
#sol
#rollingcount
#trident
#uniquevisitor
#pageview
#grep
#dataclean
#drpc- Run.
benchmark/bin/run-all.sh- Check. The benchmark results will be stored at config path METRICS_PATH(default is: /root/benchmark/reports). It contains througput data and latency of the whole cluster.
The result of wordcount contains two files
1. `WordCount_metrics_1402148415021.csv`, performnace data.
2. `WordCount_metrics_1402148415021.yaml`. The config used to run this test.
Please contact:
- Manu Zhang: tianlun.zhang@intel.com
- Sean Zhong: xiang.zhong@intel.com
We use the SOL benchmark code(https://github.com/yahoo/storm-perf-test) from yahoo. Thanks.