<<<#———————————————————————————#———————————————————————————#———————————————————————————>>> 一、tpcc <<<#———————————————————————————#———————————————————————————#———————————————————————————>>> 工具.0 /xxx/xxx/benchmarksql-5.0/run/probs.ob 建表脚本 参数:4.2.1TPCC-cs.md https://www.oceanbase.com/docs/community-tutorials-cn-10000000000012261 <<<#———————————————————————————#———————————————————————————#>>> 1、准备工具 解压benchmarksql-5.0 修改probs.ob文件 warehouses=xxx 多少仓数据,1000仓=100g,但是ob有数据压缩 loadWorkers= (案例写的20) terminals=100 并发 runMins=60 时间 osCollectorDevices=net_xxx blk_sda 网络相关 —xxx = ip addr 当前ip的网卡名 2、造数 ./runLoader.sh probs.ob 然后select count() 各表 select /+ parallel(32) / count() from bmsql_config select /+ parallel(32) / count() from bmsql_warehouse; select /+ parallel(32) / count() from bmsql_district; select /+ parallel(32) / count() from bmsql_customer; select /+ parallel(32) / count() from bmsql_history; select /+ parallel(32) / count() from bmsql_new_order; select /+ parallel(32) / count() from bmsql_oorder; select /+ parallel(32) / count() from bmsql_order_line; select /+ parallel(32) / count() from bmsql_item; select /*+ parallel(32) / count() from bmsql_stock;

3、查看到ob的大小 select svr_ip,sum(data_size)/1024/1024/1024 from dba_ob_tablet_replicas group by svr_ip;

4、执行 ./runBenchmark.sh probs.ob

<<<#———————————————————————————#———————————————————————————#———————————————————————————>>> 二、tpch <<<#———————————————————————————#———————————————————————————#———————————————————————————>>> 工具.0.0 /xxx/TPC-H_Tools_v3.0.0/dbgen 建表脚本 参数 :4.2.1TPCC-cs.md —这俩都参考一下吧 sql:D:\zcj\自我笔记\ob\qt\任务\TPCC+H\tpch\sql
https://www.oceanbase.com/docs/community-tutorials-cn-10000000000012262 <<<#———————————————————————————#———————————————————————————#>>> 1、准备工具 cd /xxx/TPC-H_Tools_v3.0.0/dbgen cp makefile.suite makefile vi makefile CC=gcc DATABASE=ORACLE MARCHINE=LINUX WORKLOAD=TPCH make -f makefile 2、 (选做,如果你是mysql,你需要在这里面加新的一条) /xxx/TPC-H_Tools_v3.0.0/dbgen/tpcd.h #ifdef MYSQL #define GEN_QUERY_PLAN "" #define START_TRAN “START TRANSACTION” #define END_TRAN “COMMIT” #define SET_OUTPUT "" #define SET_ROWCOUNT “limit %d;\n” #define SET_DBASE “use %s;\n” #endif

3、准备数据 mkdir /obbackup/db1 cp dists.dss /obbackup/db1 cp dbgen /obbackup/db1

./dbgen -s 100 -f -s 100代表100g -f代表覆盖

—选做(如果你想3个节点一起跑,这样造数快,需要后续脚本配置) 3、① mkdir /obbackup/db2 cp dists.dss /obbackup/db2 cp dbgen /obbackup/db2

1
mkdir /obbackup/db3
2
cp dists.dss /obbackup/db3
3
cp dbgen /obbackup/db3
4

5
3、②vi gen_data.sh
6
  --也是三份
7
split=1
8
END_SPLIT=100
9
while [ $split -le $END_SPLIT]
10
do
11
 ./dbgen -f -s xxxx -C300 -S $split 2>&1 &
12
 split =$[ split + 1]
13
done
14

15
split=101
16
END_SPLIT=200
17
while [ $split -le $END_SPLIT]
18
do
19
 ./dbgen -f -s xxxx -C300 -S $split 2>&1 &
20
 split =$[ split + 1]
21
done
22

23
split=201
24
END_SPLIT=300
25
while [ $split -le $END_SPLIT]
26
do
27
 ./dbgen -f -s xxxx -C300 -S $split 2>&1 &
28
 split =$[ split + 1]
29
done

4、准备库: create user tpch identified by aaAA11__; grant dba to tpch;

执行TPC-H脚本的内容, (选做:清理↓) alter table lineitem set tablegroup ”; alter table orders set tablegroup ”; alter table partsupp set tablegroup ”; alter table part set tablegroup ”; alter table customer set tablegroup ”; alter table supplier set tablegroup ”; alter table nation set tablegroup ”; alter table region set tablegroup ”;

drop table group tpch_tg_1000g_lineitem_order_group; drop table group tpch_tg_1000g_partsupp_part;

5、配置路径权限: obclient -S /home/admin/oceanbase/run/sql.sock -utpch@ -p’aaAA11__’ select * from sys.tenant_virtual_global_variable where variable_name like ‘%secure%’; set global secure_file_priv=‘/obbackup/db1/’

6、加载数据 set @@session.ob_query_timeout = 7200 000000;

load data /+ load_batch_size(1000) parallel(64)/ infile ‘/obbackup/db1/customer.tbl’ into table customer fields terminated by ’|’; load data /+ load_batch_size(1000) parallel(64)/ infile ‘/obbackup/db1/lineitem.tbl’ into table lineitem fields terminated by ’|’; load data /+ load_batch_size(1000) parallel(64)/ infile ‘/obbackup/db1/nation.tbl’ into table nation fields terminated by ’|’; load data /+ load_batch_size(1000) parallel(64)/ infile ‘/obbackup/db1/orders.tbl’ into table orders fields terminated by ’|’; load data /+ load_batch_size(1000) parallel(64)/ infile ‘/obbackup/db1/partsupp.tbl’ into table partsupp fields terminated by ’|’; load data /+ load_batch_size(1000) parallel(64)/ infile ‘/obbackup/db1/part.tbl’ into table part fields terminated by ’|’; load data /+ load_batch_size(1000) parallel(64)/ infile ‘/obbackup/db1/region.tbl’ into table region fields terminated by ’|’; load data /+ load_batch_size(1000) parallel(64)/ infile ‘/obbackup/db1/supplier.tbl’ into table supplier fields terminated by ’|’;

select /+ parallel(32) / count() from lineitem ; select /+ parallel(32) / count() from orders ; select /+ parallel(32) / count() from partsupp ; select /+ parallel(32) / count() from part ; select /+ parallel(32) / count() from customer ; select /+ parallel(32) / count() from supplier ; select /+ parallel(32) / count() from nation ; select /+ parallel(32) / count() from region ;

6、执行22个sql并记录 D:\zcj\自我笔记\ob\qt\任务\TPCC+H\tpch\sql\