HBase Cassandra Benchmark

Overview

This benchmark is designed to run with HBase and Cassandra and automates the steps in benchmark to make sure the result is reproducible

Pre-requisites

HBase or Cassandra installed. The benchmark is designed to run on dedicated cluster.
Dedicated hardware. AWS can have noisy neighbours and thus is not a reliable place for benchmark, unless you are using dedicated instances.
Maven. YCSB depends on maven to run
YCSB Repo here

Command steps

Clone YCSB and download code from this repo, place into the same YCSB root folder. So you should see bench.sh under your YCSB folder

HBase steps

Place hbase-site.xml (Usually found under /etc/hbase/hbase.conf/hbase-site.xml) into ./YCSB/hbase10/conf

Note: You'll likely have to create the conf directory by yourself.

In HBase shell ($>hbase shell) run the follwoing command to create tables

create 'usertable', 'cf', {SPLITS => (1..200).map {|i| "user#{1000+i*(9999-1000)/200}"}, MAX_FILESIZE => 
4*1024**3}

Run ./bench_hbase.sh to start benchmark Results will be logged in individual log files

For more details, see here: https://github.com/brianfrankcooper/YCSB/tree/master/hbase10

Cassandra steps

Open bench_cassandra.sh, edit "hosts" (line 9) variable. Replace it with your cassandra instance's IP or hostname

Run cqlsh host to get into cql shell. Then run

cqlsh> create keyspace ycsb
    WITH REPLICATION = {'class' : 'SimpleStrategy', 'replication_factor': 3 };
cqlsh> USE ycsb;
cqlsh> create table usertable (
    y_id varchar primary key,
    field0 varchar,
    field1 varchar,
    field2 varchar,
    field3 varchar,
    field4 varchar,
    field5 varchar,
    field6 varchar,
    field7 varchar,
    field8 varchar,
    field9 varchar);

Run ./bench_cassandra.sh to start benchmark Results will be logged in individual log files.

For more details, see here: https://github.com/brianfrankcooper/YCSB/tree/master/cassandra

Questions?

Reach out to me via message or leave a comment. I'll help you out!