YCSB测试HBase的使用方法

n_splits = 20
create 'usertable','family',{SPLITS => (1..n_splits).map {|i| "user#{1000+i*(9999-1000)/n_splits}"}}

bin/ycsb run hbase10 -P workloads/workloada -cp /var/rds/hbase-1.3.0/conf -p table=usertable -p columnfamily=family -threads 90

bin/ycsb run hbase10 -P workloads/workloadc -cp /opt/tools/hbase-1.3.0/conf -p table=usertable -p columnfamily=family -threads 90 -p maxexecutiontime=1800 -s > workloadc-result.txt

bin/ycsb run hbase10 -P workloads/workloadx -cp /opt/tools/hbase-1.3.0/conf -p table=usertable -p columnfamily=family -threads 90 -p maxexecutiontime=1800 -s 1> workloadx-run3.txt -s 2> workloadx-result3.txt

测试读操作前，需要先load数据，才能在读操作中被识别
bin/ycsb load hbase10 -P workloads/workloada -cp /var/rds/hbase-1.3.0/conf -p table=usertable -p columnfamily=family -threads 90 -p maxexecutiontime=1800 -s 1> workloada-run1.txt -s 2> workloada-result1.txt

获取OPS
cat current workloadx-result4.txt | awk '{print $7}'

P99时延
cat current workloadx-result4.txt | awk '{print $21}' | awk -F '=' '{print $2}' | awk -F ',' '{print $1}'

平均时延
cat current workloadx-result4.txt | awk '{print $19}' | awk -F '=' '{print $2}' | awk -F ',' '{print $1}'