Oceanbase使用BenchmarkSQL 进行tpcc测试及执行计划查询
文章目录
- Oceanbase使用BenchmarkSQL 进行tpcc测试及执行计划查询
- 环境介绍
- 配置BenchmarkSQL
- 开始BenchmarkSQL测试
- 查看执行计划
- 查询top sql
- 查看执行计划
- 总结
- 参考文档
环境介绍
操作系统：redhat7.9
内核版本：3.10.0-1160.el7.x86_64
Ocaenbase集群版本：社区版单副本集群，3.1.0，租户为mysql模式
Benchmarksql要求jdk8版本及以上的java环境
[admin@rac04 ~]$ java -version
openjdk version "1.8.0_262"
OpenJDK Runtime Environment (build 1.8.0_262-b10)
OpenJDK 64-Bit Server VM (build 25.262-b10, mixed mode)
配置BenchmarkSQL
创建ob的tpcc数据库和tpcc用户
create database tpcc;
create user tpcc identified by 'tpcc';
grant all on *.* to tpcc;
修改ob的超时设置
set global ob_query_timeout=36000000000;
set global ob_trx_timeout=36000000000;-- 为了获得top sql，需要将sql审计打开，默认是开启的
show variables like 'ob_enable_sql_audit';
MySQL [tpcc]> show variables like 'ob_enable_sql_audit';
+---------------------+-------+
| Variable_name | Value |
+---------------------+-------+
| ob_enable_sql_audit | ON |
+---------------------+-------+
下载BenchmarkSQL
su - admin
wget -O benchmarksql-5.0.zip https://github.com/meiq4096/benchmarksql-5.0/archive/refs/heads/master.zip
配置benchmarksql
unzip benchmarksql-5.0.zip
cd benchmarksql-5.0-master/run
# 编辑props.ob配置文件
cp props.ob props.ob.bak20240312
vi props.obdb=oceanbase
driver=com.alipay.oceanbase.obproxy.mysql.jdbc.Driver
conn=jdbc:oceanbase://127.1:2883/tpcc?useUnicode=true&characterEncoding=utf-8
user=tpcc@yw#obcluster
password=tpcc// 仓库数为2
warehouses=2
// 并发数为5
loadWorkers=5terminals=2
//To run specified transactions per terminal- runMins must equal zero
runTxnsPerTerminal=0
//To run for specified minutes- runTxnsPerTerminal must equal zero
// 运行10分钟
runMins=10
//Number of total transactions per minute
limitTxnsPerMin=0//Set to true to run in 4.x compatible mode. Set to false to use the
//entire configured database evenly.
terminalWarehouseFixed=true//The following five values must add up to 100
newOrderWeight=45
paymentWeight=43
orderStatusWeight=4
deliveryWeight=4
stockLevelWeight=4// Directory name to create for collecting detailed result data.
// Comment this out to suppress.
resultDirectory=my_result_%tY-%tm-%td_%tH%tM%tS
osCollectorScript=./misc/os_collector_linux.py
osCollectorInterval=1
修改创建表的脚本
cd /sql.oceanbase
cp tableCreates.sql tableCreates.sql.bak20240312
# varchar2类型字段，与mysql租户不兼容，修改varchar2为varchar
sed -i 's/varchar2/varchar/g' tableCreates.sql
# 脚本中创建表组的语句有误，表组名称不能使用双引号
sed -i 's/\"tpcc_group\"/tpcc_group/g' tableCreates.sql
# mysql类型的租户不支持sequence，删除创建sequence的语句
sed -i 's/create sequence bmsql_hist_id_seq;//g' tableCreates.sql
cat tableCreates.sql
创建表
[admin@rac04 run]$ pwd
/software/benchmarksql-5.0-master/run
[admin@rac04 run]$ ./runSQL.sh props.ob sql.oceanbase/tableCreates.sql
校验表
MySQL [tpcc]> show tables;
+------------------+
| Tables_in_tpcc |
+------------------+
| bmsql_config |
| bmsql_customer |
| bmsql_district |
| bmsql_history |
| bmsql_item |
| bmsql_new_order |
| bmsql_oorder |
| bmsql_order_line |
| bmsql_stock |
| bmsql_warehouse |
+------------------+
10 rows in set (0.004 sec)
加载数据入表
[admin@rac04 run]$ ./runLoader.sh props.ob
Starting BenchmarkSQL LoadDatadriver=com.alipay.oceanbase.obproxy.mysql.jdbc.Driver
conn=jdbc:oceanbase://127.1:2883/tpcc?useUnicode=true&characterEncoding=utf-8
user=tpcc@yw#obcluster
password=***********
warehouses=2
loadWorkers=5
fileLocation (not defined)
csvNullValue (not defined - using default 'NULL')Worker 000: Loading ITEM
Worker 001: Loading Warehouse 1
Worker 002: Loading Warehouse 2
Worker 000: Loading ITEM done
Worker 001: Loading Warehouse 1 done
Worker 002: Loading Warehouse 2 done
创建索引
[admin@rac04 run]$ ./runSQL.sh props.ob sql.oceanbase/indexCreates.sql
# ------------------------------------------------------------
# Loading SQL file sql.oceanbase/indexCreates.sql
# ------------------------------------------------------------
create index bmsql_customer_idx1
on bmsql_customer (c_w_id, c_d_id, c_last, c_first) local;
create index bmsql_oorder_idx1
on bmsql_oorder (o_w_id, o_d_id, o_carrier_id, o_id) local;
开始BenchmarkSQL测试
开始测试之前，手动做一次合并
MySQL [oceanbase]> alter system major freeze;
Query OK, 0 rows affected (0.007 sec)
开始测试
[admin@rac04 run]$ ./runBenchmark.sh props.ob
01:27:24,021 [main] INFO jTPCC : Term-00,
01:27:24,028 [main] INFO jTPCC : Term-00, +-------------------------------------------------------------+
01:27:24,029 [main] INFO jTPCC : Term-00, BenchmarkSQL v5.0
01:27:24,029 [main] INFO jTPCC : Term-00, +-------------------------------------------------------------+
01:27:24,029 [main] INFO jTPCC : Term-00, (c) 2003, Raul Barbosa
01:27:24,029 [main] INFO jTPCC : Term-00, (c) 2004-2016, Denis Lussier
01:27:24,031 [main] INFO jTPCC : Term-00, (c) 2016, Jan Wieck
01:27:24,031 [main] INFO jTPCC : Term-00, +-------------------------------------------------------------+
01:27:24,031 [main] INFO jTPCC : Term-00,
01:27:24,031 [main] INFO jTPCC : Term-00, db=oracle
01:27:24,032 [main] INFO jTPCC : Term-00, driver=com.alipay.oceanbase.obproxy.mysql.jdbc.Driver
01:27:24,032 [main] INFO jTPCC : Term-00, conn=jdbc:oceanbase://127.1:2883/tpcc?useUnicode=true&characterEncoding=utf-8
01:27:24,032 [main] INFO jTPCC : Term-00, user=tpcc@yw#obcluster
01:27:24,032 [main] INFO jTPCC : Term-00,
01:27:24,032 [main] INFO jTPCC : Term-00, warehouses=2
01:27:24,032 [main] INFO jTPCC : Term-00, terminals=2
01:27:24,033 [main] INFO jTPCC : Term-00, runMins=10
01:27:24,033 [main] INFO jTPCC : Term-00, limitTxnsPerMin=0
01:27:24,034 [main] INFO jTPCC : Term-00, terminalWarehouseFixed=true
01:27:24,034 [main] INFO jTPCC : Term-00,
01:27:24,034 [main] INFO jTPCC : Term-00, newOrderWeight=45
01:27:24,034 [main] INFO jTPCC : Term-00, paymentWeight=43
01:27:24,034 [main] INFO jTPCC : Term-00, orderStatusWeight=4
01:27:24,034 [main] INFO jTPCC : Term-00, deliveryWeight=4
01:27:24,034 [main] INFO jTPCC : Term-00, stockLevelWeight=4
01:27:24,034 [main] INFO jTPCC : Term-00,
01:27:24,034 [main] INFO jTPCC : Term-00, resultDirectory=my_result_%tY-%tm-%td_%tH%tM%tS
01:27:24,034 [main] INFO jTPCC : Term-00, osCollectorScript=./misc/os_collector_linux.py
01:27:24,034 [main] INFO jTPCC : Term-00,
01:27:24,045 [main] INFO jTPCC : Term-00, copied props.ob to my_result_2024-03-12_012724/run.properties
01:27:24,045 [main] INFO jTPCC : Term-00, created my_result_2024-03-12_012724/data/runInfo.csv for runID 6
01:27:24,046 [main] INFO jTPCC : Term-00, writing per transaction results to my_result_2024-03-12_012724/data/result.csv
01:27:24,047 [main] INFO jTPCC : Term-00, osCollectorScript=./misc/os_collector_linux.py
01:27:24,047 [main] INFO jTPCC : Term-00, osCollectorInterval=1
01:27:24,047 [main] INFO jTPCC : Term-00, osCollectorSSHAddr=null
01:27:24,047 [main] INFO jTPCC : Term-00, osCollectorDevices=null
01:27:24,134 [main] INFO jTPCC : Term-00,
01:27:24,420 [main] INFO jTPCC : Term-00, C value for C_LAST during load: 19
01:27:24,421 [main] INFO jTPCC : Term-00, C value for C_LAST this run: 122
01:27:24,421 [main] INFO jTPCC : Term-00,
……
……Term-00, Running Average tpmTOTAL: 1434.69 Current tpmTOTAL: 95832 Memory Usage: 18MB / 196MB
01:37:24,679 [Thread-1] INFO jTPCC : Term-00,
01:37:24,679 [Thread-1] INFO jTPCC : Term-00,
01:37:24,679 [Thread-1] INFO jTPCC : Term-00, Measured tpmC (NewOrders) = 637.1
01:37:24,680 [Thread-1] INFO jTPCC : Term-00, Measured tpmTOTAL = 1434.47
01:37:24,680 [Thread-1] INFO jTPCC : Term-00, Session Start = 2024-03-12 01:27:24
01:37:24,680 [Thread-1] INFO jTPCC : Term-00, Session End = 2024-03-12 01:37:24
01:37:24,680 [Thread-1] INFO jTPCC : Term-00, Transaction Count = 14348
查看执行计划
查询top sql
查找性能测试期间，执行时间top5 sql
这里着重解释一下request_time的条件，以性能测试结束时间为基点，查找请求时间位于2024-03-12 01:37:24及之前10分钟内的sql。需要注意的是将request_time的值是距离 1970-01-01 00:00:00.000000 的微秒数，因此需要调用time_to_usec函数转换时间。
SELECT/*+ PARALLEL(4)*/ sql_id,query_sql,plan_id, elapsed_time , trace_id FROM oceanbase.gv$sql_audit WHERE tenant_id = 1001 and IS_EXECUTOR_RPC = 0 and request_time > (time_to_usec(date_format('2024-03-12 01:37:24','%Y-%m-%d %T')) - 10000000*6*10) AND request_time < time_to_usec(date_format('2024-03-12 01:37:24','%Y-%m-%d %T')) ORDER BY elapsed_time DESC LIMIT 10;MySQL [tpcc]> SELECT/*+ PARALLEL(4)*/ sql_id,query_sql,plan_id, elapsed_time , trace_id -> FROM oceanbase.gv$sql_audit -> WHERE tenant_id = 1001 -> and IS_EXECUTOR_RPC = 0 -> and request_time > (time_to_usec(date_format('2024-03-12 01:37:24','%Y-%m-%d %T')) - 10000000*6*10) -> AND request_time < time_to_usec(date_format('2024-03-12 01:37:24','%Y-%m-%d %T')) -> ORDER BY elapsed_time DESC LIMIT 10\G
*************************** 1. row ***************************sql_id: FFFCA4D67EA0A788813031B8BBC3B329query_sql: SELECT count(*) AS low_stock FROM ( SELECT s_w_id, s_i_id, s_quantity FROM bmsql_stock WHERE s_w_id = 2 AND s_quantity < 12 AND s_i_id IN ( SELECT ol_i_id FROM bmsql_district JOIN bmsql_order_line ON ol_w_id = d_w_id AND ol_d_id = d_id AND ol_o_id >= d_next_o_id - 20 AND ol_o_id < d_next_o_id WHERE d_w_id = 2 AND d_id = 5 ) )plan_id: 130
elapsed_time: 19786trace_id: YB420B0B0B0E-00061364401FFA67
*************************** 2. row ***************************sql_id: FFFCA4D67EA0A788813031B8BBC3B329query_sql: SELECT count(*) AS low_stock FROM ( SELECT s_w_id, s_i_id, s_quantity FROM bmsql_stock WHERE s_w_id = 1 AND s_quantity < 13 AND s_i_id IN ( SELECT ol_i_id FROM bmsql_district JOIN bmsql_order_line ON ol_w_id = d_w_id AND ol_d_id = d_id AND ol_o_id >= d_next_o_id - 20 AND ol_o_id < d_next_o_id WHERE d_w_id = 1 AND d_id = 10 ) )plan_id: 130
elapsed_time: 18749trace_id: YB420B0B0B0E-000613643FFFF583
*************************** 3. row ***************************sql_id: FFFCA4D67EA0A788813031B8BBC3B329query_sql: SELECT count(*) AS low_stock FROM ( SELECT s_w_id, s_i_id, s_quantity FROM bmsql_stock WHERE s_w_id = 1 AND s_quantity < 16 AND s_i_id IN ( SELECT ol_i_id FROM bmsql_district JOIN bmsql_order_line ON ol_w_id = d_w_id AND ol_d_id = d_id AND ol_o_id >= d_next_o_id - 20 AND ol_o_id < d_next_o_id WHERE d_w_id = 1 AND d_id = 10 ) )plan_id: 130
elapsed_time: 18074trace_id: YB420B0B0B0E-00061364402FC346
*************************** 4. row ***************************sql_id: FFFCA4D67EA0A788813031B8BBC3B329query_sql: SELECT s_quantity, s_data, s_dist_01, s_dist_02, s_dist_03, s_dist_04, s_dist_05, s_dist_06, s_dist_07, s_dist_08, s_dist_09, s_dist_10 FROM bmsql_stock WHERE s_w_id = 2 AND s_i_id = 68621 FOR UPDATEplan_id: 116
elapsed_time: 17701trace_id: YB420B0B0B0E-00061364404FC6E5
*************************** 5. row ***************************sql_id: FFFCA4D67EA0A788813031B8BBC3B329query_sql: SELECT s_quantity, s_data, s_dist_01, s_dist_02, s_dist_03, s_dist_04, s_dist_05, s_dist_06, s_dist_07, s_dist_08, s_dist_09, s_dist_10 FROM bmsql_stock WHERE s_w_id = 2 AND s_i_id = 59773 FOR UPDATEplan_id: 116
elapsed_time: 16933trace_id: YB420B0B0B0E-00061364400FF6EF
*************************** 6. row ***************************sql_id: query_sql: SELECT s_quantity, s_data, s_dist_01, s_dist_02, s_dist_03, s_dist_04, s_dist_05, s_dist_06, s_dist_07, s_dist_08, s_dist_09, s_dist_10 FROM bmsql_stock WHERE s_w_id = 2 AND s_i_id = 45283 FOR UPDATEplan_id: 116
elapsed_time: 16895trace_id: YB420B0B0B0E-00061364401FFAE2
*************************** 7. row ***************************sql_id: FFFCA4D67EA0A788813031B8BBC3B329query_sql: SELECT s_quantity, s_data, s_dist_01, s_dist_02, s_dist_03, s_dist_04, s_dist_05, s_dist_06, s_dist_07, s_dist_08, s_dist_09, s_dist_10 FROM bmsql_stock WHERE s_w_id = 2 AND s_i_id = 83525 FOR UPDATEplan_id: 116
elapsed_time: 15588trace_id: YB420B0B0B0E-00061364401FF927
*************************** 8. row ***************************sql_id: FFFCA4D67EA0A788813031B8BBC3B329query_sql: SELECT i_price, i_name, i_data FROM bmsql_item WHERE i_id = 39891plan_id: 113
elapsed_time: 15104trace_id: YB420B0B0B0E-00061364403FC65B
*************************** 9. row ***************************sql_id: FFFCA4D67EA0A788813031B8BBC3B329query_sql: SELECT c_first, c_middle, c_last, c_street_1, c_street_2, c_city, c_state, c_zip, c_phone, c_since, c_credit, c_credit_lim, c_discount, c_balance FROM bmsql_customer WHERE c_w_id = 2 AND c_d_id = 4 AND c_id = 193 FOR UPDATEplan_id: 118
elapsed_time: 15015trace_id: YB420B0B0B0E-00061364401FF7C1
*************************** 10. row ***************************sql_id: FFFCA4D67EA0A788813031B8BBC3B329query_sql: SELECT i_price, i_name, i_data FROM bmsql_item WHERE i_id = 4421plan_id: 113
elapsed_time: 14234trace_id: YB420B0B0B0E-00061364400FFABC
10 rows in set (0.026 sec)
查看执行计划
首先查看top10的解析执行计划，这里挑3条，plan_id分别为130，116，113。
MySQL [tpcc]> explain SELECT count(*) AS low_stock FROM ( SELECT s_w_id, s_i_id, s_quantity FROM bmsql_stock WHERE s_w_id = 2 AND s_quantity < 12 AND s_i_id IN ( SELECT ol_i_id FROM bmsql_district JOIN bmsql_order_line ON ol_w_id = d_w_id AND ol_d_id = d_id AND ol_o_id >= d_next_o_id - 20 AND ol_o_id < d_next_o_id WHERE d_w_id = 2 AND d_id = 5));
+----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
| Query Plan |
+----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
| ==========================================================
|ID|OPERATOR |NAME |EST. ROWS|COST |
----------------------------------------------------------
|0 |SCALAR GROUP BY | |1 |46984|
|1 | NESTED-LOOP JOIN | |1221 |46750|
|2 | SUBPLAN SCAN |VIEW2 |1215 |2677 |
|3 | HASH DISTINCT | |1215 |2509 |
|4 | NESTED-LOOP JOIN| |1219 |1239 |
|5 | TABLE GET |bmsql_district |1 |53 |
|6 | TABLE SCAN |bmsql_order_line|3656 |1194 |
|7 | TABLE GET |bmsql_stock |1 |37 |
==========================================================Outputs & filters:
-------------------------------------0 - output([T_FUN_COUNT(*)]), filter(nil), group(nil), agg_func([T_FUN_COUNT(*)])1 - output([1]), filter(nil), conds(nil), nl_params_([VIEW2.VIEW1.ol_i_id])2 - output([VIEW2.VIEW1.ol_i_id]), filter(nil), access([VIEW2.VIEW1.ol_i_id])3 - output([bmsql_order_line.ol_i_id]), filter(nil), distinct([bmsql_order_line.ol_i_id])4 - output([bmsql_order_line.ol_i_id]), filter(nil), conds(nil), nl_params_([bmsql_district.d_next_o_id - 20], [bmsql_district.d_next_o_id])5 - output([bmsql_district.d_next_o_id]), filter([bmsql_district.d_next_o_id > bmsql_district.d_next_o_id - 20]), access([bmsql_district.d_next_o_id]), partitions(p0)6 - output([bmsql_order_line.ol_i_id]), filter(nil), access([bmsql_order_line.ol_i_id]), partitions(p0)7 - output([1]), filter([bmsql_stock.s_quantity < 12]), access([bmsql_stock.s_quantity]), partitions(p0)|
+----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
1 row in set (0.010 sec)
MySQL [tpcc]> explain SELECT s_quantity, s_data, s_dist_01, s_dist_02, s_dist_03, s_dist_04, s_dist_05, s_dist_06, s_dist_07, s_dist_08, s_dist_09, s_dist_10 FROM bmsql_stock WHERE s_w_id = 2 AND s_i_id = 59773 FOR UPDATE;
+--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
| Query Plan |
+--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
| =========================================
|ID|OPERATOR |NAME |EST. ROWS|COST|
-----------------------------------------
|0 |TABLE GET|bmsql_stock|1 |54 |
=========================================Outputs & filters:
-------------------------------------0 - output([bmsql_stock.s_quantity], [bmsql_stock.s_data], [bmsql_stock.s_dist_01], [bmsql_stock.s_dist_02], [bmsql_stock.s_dist_03], [bmsql_stock.s_dist_04], [bmsql_stock.s_dist_05], [bmsql_stock.s_dist_06], [bmsql_stock.s_dist_07], [bmsql_stock.s_dist_08], [bmsql_stock.s_dist_09], [bmsql_stock.s_dist_10]), filter(nil), access([bmsql_stock.s_w_id], [bmsql_stock.s_i_id], [bmsql_stock.s_quantity], [bmsql_stock.s_data], [bmsql_stock.s_dist_01], [bmsql_stock.s_dist_02], [bmsql_stock.s_dist_03], [bmsql_stock.s_dist_04], [bmsql_stock.s_dist_05], [bmsql_stock.s_dist_06], [bmsql_stock.s_dist_07], [bmsql_stock.s_dist_08], [bmsql_stock.s_dist_09], [bmsql_stock.s_dist_10]), partitions(p0)|
+--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
1 row in set (0.005 sec)
MySQL [tpcc]> explain SELECT i_price, i_name, i_data FROM bmsql_item WHERE i_id = 39891;
+-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
| Query Plan |
+-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
| ========================================
|ID|OPERATOR |NAME |EST. ROWS|COST|
----------------------------------------
|0 |TABLE GET|bmsql_item|1 |53 |
========================================Outputs & filters:
-------------------------------------0 - output([bmsql_item.i_price], [bmsql_item.i_name], [bmsql_item.i_data]), filter(nil), access([bmsql_item.i_price], [bmsql_item.i_name], [bmsql_item.i_data]), partitions(p0)|
+-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
1 row in set (0.004 sec)
查看top3的实际执行计划。由于执行计划已经刷出缓存，因此性能测试期间top10sql的实际执行计划没法看到。这里手动执行相关sql，再查询其执行计划。
MySQL [tpcc]> SELECT count(*) AS low_stock FROM ( SELECT s_w_id, s_i_id, s_quantity FROM bmsql_stock WHERE s_w_id = 2 AND s_quantity < 12 AND s_i_id IN ( SELECT ol_i_id FROM bmsql_district JOIN bmsql_order_line ON ol_w_id = d_w_id AND ol_d_id = d_id AND ol_o_id >= d_next_o_id - 20 AND ol_o_id < d_next_o_id WHERE d_w_id = 2 AND d_id = 5));
-- 查询刚刚执行的sql的plan_id
MySQL [tpcc1]> SELECT/*+ PARALLEL(4)*/ tenant_id,svr_ip,svr_port,sql_id,query_sql,plan_id, elapsed_time , trace_id -> FROM oceanbase.gv$sql_audit -> WHERE tenant_id = 1001 and user_name='tpcc'-> and IS_EXECUTOR_RPC = 0 -> and request_time > (time_to_usec(now()) - 10000000*6*10) -> AND request_time < time_to_usec(now()) -> ORDER BY elapsed_time DESC LIMIT 10\G
*************************** 1. row ***************************tenant_id: 1001svr_ip: 11.11.11.14svr_port: 2882sql_id: 9EDED4BC6305E5F4FCF86A6366BDD638query_sql: SELECT count(*) AS low_stock FROM ( SELECT s_w_id, s_i_id, s_quantity FROM bmsql_stock WHERE s_w_id = 2 AND s_quantity < 12 AND s_i_id IN ( SELECT ol_i_id FROM bmsql_district JOIN bmsql_order_line ON ol_w_id = d_w_id AND ol_d_id = d_id AND ol_o_id >= d_next_o_id - 20 AND ol_o_id < d_next_o_id WHERE d_w_id = 2 AND d_id = 5))plan_id: 273
elapsed_time: 21739trace_id: YB420B0B0B0E-00061364403FC7B5-- 根据sql_id查看实际执行计划
MySQL [tpcc1]> select * from oceanbase.gv$plan_cache_plan_explain where tenant_id=1001 and ip='11.11.11.14' and port='2882' and plan_id =273;
+-----------+-------------+------+---------+------------+--------------+--------------------------+------------------+------+-------+--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
| TENANT_ID | IP | PORT | PLAN_ID | PLAN_DEPTH | PLAN_LINE_ID | OPERATOR | NAME | ROWS | COST | PROPERTY |
+-----------+-------------+------+---------+------------+--------------+--------------------------+------------------+------+-------+--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
| 1001 | 11.11.11.14 | 2882 | 273 | 0 | 0 | PHY_SCALAR_AGGREGATE | NULL | 1 | 46983 | NULL |
| 1001 | 11.11.11.14 | 2882 | 273 | 1 | 1 | PHY_NESTED_LOOP_JOIN | NULL | 1221 | 46749 | NULL |
| 1001 | 11.11.11.14 | 2882 | 273 | 2 | 2 | PHY_SUBPLAN_SCAN | NULL | 1215 | 2676 | NULL |
| 1001 | 11.11.11.14 | 2882 | 273 | 3 | 3 | PHY_HASH_DISTINCT | NULL | 1215 | 2508 | NULL |
| 1001 | 11.11.11.14 | 2882 | 273 | 4 | 4 | PHY_NESTED_LOOP_JOIN | NULL | 1219 | 1238 | NULL |
| 1001 | 11.11.11.14 | 2882 | 273 | 5 | 5 | PHY_TABLE_SCAN | bmsql_district | 1 | 52 | table_rows:20, physical_range_rows:1, logical_range_rows:1, index_back_rows:0, output_rows:0, est_method:local_storage |
| 1001 | 11.11.11.14 | 2882 | 273 | 5 | 6 | PHY_TABLE_SCAN | bmsql_order_line | 3656 | 1193 | table_rows:663235, physical_range_rows:3655, logical_range_rows:3655, index_back_rows:0, output_rows:3655, est_method:local_storage, avaiable_index_name[bmsql_order_line], estimation info[table_id:1100611139453804, (table_type:1, version:0-1710180010831939-1710180010831939, logical_rc:32898, physical_rc:32898), (table_type:7, version:1710180000826357-1710180010831939-1710180031319922, logical_rc:0, physical_rc:0), (table_type:5, version:1710180000826357-1710180010831939-1710180031319922, logical_rc:0, physical_rc:0), (table_type:0, version:1710180031319922-1710180031319922-9223372036854775807, logical_rc:0, physical_rc:0)] |
| 1001 | 11.11.11.14 | 2882 | 273 | 2 | 7 | PHY_TABLE_SCAN | bmsql_stock | 1 | 36 | table_rows:200000, physical_range_rows:1, logical_range_rows:1, index_back_rows:0, output_rows:0, est_method:local_storage, estimation info[table_id:1100611139453806, (table_type:1, version:0-1710180010831939-1710180010831939, logical_rc:200000, physical_rc:200000), (table_type:7, version:1710180000827822-1710180010831939-1710180031377423, logical_rc:0, physical_rc:0), (table_type:5, version:1710180000827822-1710180010831939-1710180031377423, logical_rc:0, physical_rc:0), (table_type:0, version:1710180031377423-1710180031377423-9223372036854775807, logical_rc:0, physical_rc:0)] |
+-----------+-------------+------+---------+------------+--------------+--------------------------+------------------+------+-------+--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
8 rows in set (0.012 sec)
另外两个sql也是同样的方法
MySQL [tpcc]> SELECT s_quantity, s_data, s_dist_01, s_dist_02, s_dist_03, s_dist_04, s_dist_05, s_dist_06, s_dist_07, s_dist_08, s_dist_09, s_dist_10 FROM bmsql_stock WHERE s_w_id = 2 AND s_i_id = 59773 FOR UPDATE;
+------------+------------------------------------------+--------------------------+--------------------------+--------------------------+--------------------------+--------------------------+--------------------------+--------------------------+--------------------------+--------------------------+--------------------------+
| s_quantity | s_data | s_dist_01 | s_dist_02 | s_dist_03 | s_dist_04 | s_dist_05 | s_dist_06 | s_dist_07 | s_dist_08 | s_dist_09 | s_dist_10 |
+------------+------------------------------------------+--------------------------+--------------------------+--------------------------+--------------------------+--------------------------+--------------------------+--------------------------+--------------------------+--------------------------+--------------------------+
| 104 | BpZzCr9yN9SC8clF5llHzwmg5ANR2drV0ZM553QR | mWO9fxl80umm3A23HwjvNySD | mL3z0rJqTApoGIIJjan5bGJd | xqELg15kkS8s6HTtS9hh4sXg | dTdPg5Jxlow6W9rR8rWoZxCH | tdRCLqTvKjzL00278CuyVXnb | DXL4EfkVLKtmppr1izzGfEKU | b3JcaKms9g1AwbXJyfBP8X3f | ECNsuU7KphEUcLmp9csAaWjv | PFSSM2fGrMSrNx7stdhSb5HS | Juk9PMbM4rgf0MbUcNRQJFkr |
+------------+------------------------------------------+--------------------------+--------------------------+--------------------------+--------------------------+--------------------------+--------------------------+--------------------------+--------------------------+--------------------------+--------------------------+
1 row in set (0.012 sec)MySQL [tpcc1]> SELECT/*+ PARALLEL(4)*/ tenant_id,svr_ip,svr_port,sql_id,query_sql,plan_id, elapsed_time , trace_id -> FROM oceanbase.gv$sql_audit -> WHERE tenant_id = 1001 and user_name='tpcc'-> and IS_EXECUTOR_RPC = 0 -> and request_time > (time_to_usec(now()) - 10000000*6*10) -> AND request_time < time_to_usec(now()) -> ORDER BY elapsed_time DESC LIMIT 10\G……
*************************** 2. row ***************************tenant_id: 1001svr_ip: 11.11.11.14svr_port: 2882sql_id: 7229213613983BC5FDA15AD11EC70D01query_sql: SELECT s_quantity, s_data, s_dist_01, s_dist_02, s_dist_03, s_dist_04, s_dist_05, s_dist_06, s_dist_07, s_dist_08, s_dist_09, s_dist_10 FROM bmsql_stock WHERE s_w_id = 2 AND s_i_id = 59773 FOR UPDATEplan_id: 285
elapsed_time: 11143trace_id: YB420B0B0B0E-00061364401FFB9AMySQL [tpcc1]> select * from oceanbase.gv$plan_cache_plan_explain where tenant_id=1001 and ip='11.11.11.14' and port='2882' and plan_id =285;
+-----------+-------------+------+---------+------------+--------------+----------------+-------------+------+------+----------------------------------------------------------------------------------------------------------------------------+
| TENANT_ID | IP | PORT | PLAN_ID | PLAN_DEPTH | PLAN_LINE_ID | OPERATOR | NAME | ROWS | COST | PROPERTY |
+-----------+-------------+------+---------+------------+--------------+----------------+-------------+------+------+----------------------------------------------------------------------------------------------------------------------------+
| 1001 | 11.11.11.14 | 2882 | 285 | 0 | 0 | PHY_TABLE_SCAN | bmsql_stock | 1 | 53 | table_rows:200000, physical_range_rows:1, logical_range_rows:1, index_back_rows:0, output_rows:1, est_method:local_storage |
+-----------+-------------+------+---------+------------+--------------+----------------+-------------+------+------+----------------------------------------------------------------------------------------------------------------------------+
1 row in set (0.008 sec)
MySQL [tpcc]> SELECT i_price, i_name, i_data FROM bmsql_item WHERE i_id = 39891;
+---------+------------------------+-----------------------------------------------+
| i_price | i_name | i_data |
+---------+------------------------+-----------------------------------------------+
| 69.54 | OMb08X4TPeHtLe428FtyEC | wAUmMP35zjmdWqzAVFPMijvh9fugn6EzuSclVnNIabBy8 |
+---------+------------------------+-----------------------------------------------+
1 row in set (0.003 sec)
MySQL [tpcc1]> SELECT/*+ PARALLEL(4)*/ tenant_id,svr_ip,svr_port,sql_id,query_sql,plan_id, elapsed_time , trace_id -> FROM oceanbase.gv$sql_audit -> WHERE tenant_id = 1001 and user_name='tpcc'-> and IS_EXECUTOR_RPC = 0 -> and request_time > (time_to_usec(now()) - 10000000*6*10) -> AND request_time < time_to_usec(now()) -> ORDER BY elapsed_time DESC LIMIT 10\G
……
*************************** 2. row ***************************tenant_id: 1001svr_ip: 11.11.11.14svr_port: 2882sql_id: A460265EC2F0763A15DD27CE9E4E2200query_sql: SELECT i_price, i_name, i_data FROM bmsql_item WHERE i_id = 39891plan_id: 286
elapsed_time: 2816trace_id: YB420B0B0B0E-00061364403FC7C2
2 rows in set (0.018 sec)
MySQL [tpcc1]> select * from oceanbase.gv$plan_cache_plan_explain where tenant_id=1001 and ip='11.11.11.14' and port='2882' and plan_id =286;
+-----------+-------------+------+---------+------------+--------------+----------------+------------+------+------+----------------------------------------------------------------------------------------------------------------------------+
| TENANT_ID | IP | PORT | PLAN_ID | PLAN_DEPTH | PLAN_LINE_ID | OPERATOR | NAME | ROWS | COST | PROPERTY |
+-----------+-------------+------+---------+------------+--------------+----------------+------------+------+------+----------------------------------------------------------------------------------------------------------------------------+
| 1001 | 11.11.11.14 | 2882 | 286 | 0 | 0 | PHY_TABLE_SCAN | bmsql_item | 1 | 52 | table_rows:100000, physical_range_rows:1, logical_range_rows:1, index_back_rows:0, output_rows:1, est_method:local_storage |
+-----------+-------------+------+---------+------------+--------------+----------------+------------+------+------+----------------------------------------------------------------------------------------------------------------------------+
1 row in set (0.008 sec)
总结
-
使用benchmarksql的create_table.sql脚本创建表之前，需要修改一些内容：
- varchar2–>varchar；
- 创建tablegroup的语句，表组名不能使用双引号，需要删除双引号；
- 删除创建sequence的语句。
-
查看执行计划的一般步骤
-
（可选）查询oceanbase.gv$sql_audit 视图获取top sql，获取tenant_id,svr_ip,svr_port,sql_id,query_sql,plan_id等关键信息
--查看10分钟内的耗时topsql SELECT/*+ PARALLEL(4)*/ tenant_id,svr_ip,svr_port,sql_id,query_sql,plan_id, elapsed_time , trace_id FROM oceanbase.gv$sql_audit WHERE tenant_id = 1001 and user_name='tpcc'and IS_EXECUTOR_RPC = 0 and request_time > (time_to_usec(now()) - 10000000*6*10) AND request_time < time_to_usec(now()) ORDER BY elapsed_time DESC LIMIT 10;-- 查看指定时间段内的耗时topsql，这里是2024-03-12 01:37:24至前10分钟 SELECT/*+ PARALLEL(4)*/ * FROM oceanbase.gv$sql_audit WHERE tenant_id = 1001 and IS_EXECUTOR_RPC = 0 and request_time > (time_to_usec(date_format('2024-03-12 01:37:24','%Y-%m-%d %T')) - 10000000*6*10) AND request_time < time_to_usec(date_format('2024-03-12 01:37:24','%Y-%m-%d %T')) ORDER BY elapsed_time DESC LIMIT 10; -
查询oceanbase.gv$plan_cache_plan_explain视图获取真实执行计划（缓存被清理就看不到了）
select* fromoceanbase.gv$plan_cache_plan_explain wheretenant_id = 1001and ip = '11.11.11.14'and port = '2882'and plan_id = 286; -
对于解析执行计划，直接使用explain命令即可，比如：
explain SELECT i_price, i_name, i_data FROM bmsql_item WHERE i_id = 39891;
-
参考文档
OceanBase分布式数据库-海量数据 笔笔算数
GitHub - meiq4096/benchmarksql-5.0: benchmarksql-5.0 for oceanbase
查询某段时间内执行时间排名 TOP N 的请求 -OceanBase 数据库-OceanBase文档中心-分布式数据库使用文档
查询排名 TOP N 的 SQL -OceanBase 数据库-OceanBase文档中心-分布式数据库使用文档
函数 -OceanBase 数据库-OceanBase文档中心-分布式数据库使用文档
 A~D)
![[Django 0-1] Core.Cache模块](http://pic.xiahunao.cn/[Django 0-1] Core.Cache模块)
