您好,登錄后才能下訂單哦!
性能分析過程中,經常會遇到生產庫出現SQL的性能問題,但是,我們沒有辦法在生產庫上做很多動作,需要將這個SQL的對應的表結構信息,統計信息導入到測試庫進行測試(沒有真實的測試數據,只有統計信息)
本次測試目的,客戶環境導出了Test和product環境的相同SQL的不同執行計劃的SQLT報告。
TEST環境是11.2.0.1版本
Product是11.2.0.2版本
我們會將兩個SQLT報告分別導入到11.2.0.1和11.2.0.2的測試環境中,然后將11.2.0.1版本,好的執行計劃,作為profile導出到11.2.0.2中。讓11.2.0.2也根據11.2.0.1中的SQL一樣,使用相同的執行計劃。
~~~~~~~第一部分,導入testcase環境,只導入統計信息,和表的元數據。沒有真實數據~~~~~~
1. 參考sqlt_s39917_readme.html(客戶SQLT輸出的文件中,就已經包含這個readme.html文件)
也可以參考文檔How to Use SQLT (SQLTXPLAIN) to Create a Testcase Without Row Data (Doc ID 1470811.1)
Implement SQLT Test Case (TC) Custom mode Create test case user and schema objects connecting as SYSDBA: sqlplus / as sysdba START sqlt_s39917_metadata.sql Purge pre-existing s39917 from local SQLT repository connected as SYSDBA: START sqlt_s39917_purge.sql Import SQLT repository for s39917 (provide SQLTXPLAIN password): HOS imp sqltxplain FILE=sqlt_s39917_exp.dmp LOG=sqlt_s39917_imp.log TABLES=sqlt% IGNORE=Y Restore CBO schema statistics for test case user connected as SYSDBA: START sqlt_s39917_restore.sql Restore CBO system statistics connected as SYSDBA: START sqlt_s39917_system_stats.sql Set the CBO environment connecting as test case user TC39917 (include optional test case user suffix): CONN TC39917/TC39917 START sqlt_s39917_set_cbo_env.sql Execute test case: START tc.sql
2. 執行tc.sql
注意:每次sqlplus連接進入后,都需要執行最后兩步
CONN TC39917/TC39917 START sqlt_s39917_set_cbo_env.sql Execute test case: START tc.sql ----需要修改q.sql中對應表的用戶名為TC39917
TC.sql執行結果如下
SQL> start tc.sql SQL> REM Executes SQL on TC then produces execution plan. Just execute "@tc.sql" from sqlplus. SQL> SET APPI OFF SERVEROUT OFF; SQL> @@q.sql SQL> SELECT Distinct fc_acc.blnctr_acc, 2 fc_acc.shortcut, 3 fc_acc.acc_name, 4 fc_subject.amount_money_sign, 5 fc_acc.subject_code, 6 fc_acc.corp_code, 7 fc_acc.net_code 8 FROM TC39917.fc_acc, TC39917.fc_user_corp, TC39917.fc_subject 9 WHERE ((fc_acc.corp_code = fc_user_corp.corp_code) OR 10 (acc_cussent is not Null or acc_cussent <> '')) 11 AND (fc_acc.subject_code = fc_subject.subject_code) 12 and ((fc_acc.checker_code is Not Null)) 13 AND (fc_acc.acc_state <> 'R') 14 AND (fc_acc.cur_code = 01) 15 and (fc_acc.acc_state <> 'D') 16 AND (fc_acc.net_code = 1000) 17 AND (fc_acc.acc_type_code = '02' or fc_subject.sys_subject_Code = '1010' or 18 (union_acc is not Null or union_acc <> '')) ; no rows selected SQL> @@plan.sql SQL> REM Displays plan for most recently executed SQL. Just execute "@plan.sql" from sqlplus. SQL> SET PAGES 2000 LIN 180; SQL> SPO plan.log; SQL> --SELECT * FROM TABLE(DBMS_XPLAN.DISPLAY_CURSOR); SQL> SELECT * FROM TABLE(DBMS_XPLAN.DISPLAY_CURSOR(NULL,NULL,'BASIC ROWS COST PREDICATE')); PLAN_TABLE_OUTPUT ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ EXPLAINED SQL STATEMENT: ------------------------ SELECT Distinct fc_acc.blnctr_acc, fc_acc.shortcut, fc_acc.acc_name, fc_subject.amount_money_sign, fc_acc.subject_code, fc_acc.corp_code, fc_acc.net_code FROM fc_acc, fc_user_corp, fc_subject WHERE ((fc_acc.corp_code = fc_user_corp.corp_code) OR (acc_cussent is not Null or acc_cussent <> '')) AND (fc_acc.subject_code = fc_subject.subject_code) and ((fc_acc.checker_code is Not Null)) AND (fc_acc.acc_state <> 'R') AND (fc_acc.cur_code = 01) and (fc_acc.acc_state <> 'D') AND (fc_acc.net_code = 1000) AND (fc_acc.acc_type_code = '02' or fc_subject.sys_subject_Code = '1010' or (union_acc is not Null or union_acc <> '')) Plan hash value: 31035937 ------------------------------------------------------------------------------ | Id | Operation | Name | Rows | Cost (%CPU)| ------------------------------------------------------------------------------ | 0 | SELECT STATEMENT | | | 133 (100)| | 1 | HASH UNIQUE | | 4573K| 133 (3)| | 2 | CONCATENATION | | | | | 3 | MERGE JOIN CARTESIAN | | 1 | 42 (0)| | 4 | NESTED LOOPS | | | | | 5 | NESTED LOOPS | | 1 | 36 (0)| |* 6 | TABLE ACCESS FULL | FC_ACC | 1 | 35 (0)| |* 7 | INDEX UNIQUE SCAN | PK_FC_SUBJECT | 1 | 0 (0)| |* 8 | TABLE ACCESS BY INDEX ROWID| FC_SUBJECT | 1 | 1 (0)| | 9 | BUFFER SORT | | 6210 | 41 (0)| | 10 | TABLE ACCESS FULL | FC_USER_CORP | 6210 | 6 (0)| | 11 | MERGE JOIN CARTESIAN | | 1 | 42 (0)| | 12 | NESTED LOOPS | | | | | 13 | NESTED LOOPS | | 1 | 36 (0)| |* 14 | TABLE ACCESS FULL | FC_ACC | 1 | 35 (0)| |* 15 | INDEX UNIQUE SCAN | PK_FC_SUBJECT | 1 | 0 (0)| |* 16 | TABLE ACCESS BY INDEX ROWID| FC_SUBJECT | 1 | 1 (0)| | 17 | BUFFER SORT | | 6210 | 41 (0)| | 18 | TABLE ACCESS FULL | FC_USER_CORP | 6210 | 6 (0)| |* 19 | HASH JOIN | | 122 | 47 (3)| |* 20 | HASH JOIN | | 7 | 41 (3)| |* 21 | TABLE ACCESS FULL | FC_ACC | 8 | 35 (0)| | 22 | TABLE ACCESS FULL | FC_SUBJECT | 636 | 5 (0)| | 23 | TABLE ACCESS FULL | FC_USER_CORP | 6210 | 6 (0)| ------------------------------------------------------------------------------ Predicate Information (identified by operation id): --------------------------------------------------- 6 - filter((TO_NUMBER("FC_ACC"."CUR_CODE")=1 AND "FC_ACC"."ACC_STATE"<>'R' AND "FC_ACC"."ACC_STATE"<>'D' AND "ACC_CUSSENT"<>'' AND "FC_ACC"."CHECKER_CODE" IS NOT NULL AND TO_NUMBER("FC_ACC"."NET_CODE")=1000)) 7 - access("FC_ACC"."SUBJECT_CODE"="FC_SUBJECT"."SUBJECT_CODE") 8 - filter(("FC_ACC"."ACC_TYPE_CODE"='02' OR "FC_SUBJECT"."SYS_SUBJECT_CODE"='1010' OR "UNION_ACC" IS NOT NULL OR "UNION_ACC"<>'')) 14 - filter(("ACC_CUSSENT" IS NOT NULL AND TO_NUMBER("FC_ACC"."CUR_CODE")=1 AND "FC_ACC"."ACC_STATE"<>'R' AND "FC_ACC"."ACC_STATE"<>'D' AND "FC_ACC"."CHECKER_CODE" IS NOT NULL AND TO_NUMBER("FC_ACC"."NET_CODE")=1000 AND LNNVL("ACC_CUSSENT"<>''))) 15 - access("FC_ACC"."SUBJECT_CODE"="FC_SUBJECT"."SUBJECT_CODE") 16 - filter(("FC_ACC"."ACC_TYPE_CODE"='02' OR "FC_SUBJECT"."SYS_SUBJECT_CODE"='1010' OR "UNION_ACC" IS NOT NULL OR "UNION_ACC"<>'')) 19 - access("FC_ACC"."CORP_CODE"="FC_USER_CORP"."CORP_CODE") 20 - access("FC_ACC"."SUBJECT_CODE"="FC_SUBJECT"."SUBJECT_CODE") filter(("FC_ACC"."ACC_TYPE_CODE"='02' OR "FC_SUBJECT"."SYS_SUBJECT_CODE"='1010' OR "UNION_ACC" IS NOT NULL OR "UNION_ACC"<>'')) 21 - filter((TO_NUMBER("FC_ACC"."CUR_CODE")=1 AND "FC_ACC"."ACC_STATE"<>'R' AND "FC_ACC"."ACC_STATE"<>'D' AND "FC_ACC"."CHECKER_CODE" IS NOT NULL AND TO_NUMBER("FC_ACC"."NET_CODE")=1000 AND LNNVL("ACC_CUSSENT" IS NOT NULL) AND LNNVL("ACC_CUSSENT"<>''))) 76 rows selected. SQL> SPO OFF; SQL>
假設,上面的執行計劃是我們期望的,性能優異的。我需要導出為profile,并導入到性能較差的數據庫
~~~~~~~~第二部分,導出11.2.0.1的profile~~~~~~~~~~~~~~~~~~~~
3. 生成profile
SQL> @/home/oracle/Bright-temp/sqlt/utl/coe_xfr_sql_profile.sql cwju3syt2mx9a 31035937 SQL> SPO coe_xfr_sql_profile.log; SQL> SET DEF ON TERM OFF ECHO ON FEED OFF VER OFF HEA ON LIN 2000 PAGES 100 LONG 8000000 LONGC 800000 TRIMS ON TI OFF TIMI OFF SERVEROUT ON SIZE 1000000 NUMF "" SQLP SQL>; Parameter 1: SQL_ID (required) PLAN_HASH_VALUE AVG_ET_SECS --------------- ----------- 31035937 .014 Parameter 2: PLAN_HASH_VALUE (required) Values passed to coe_xfr_sql_profile: ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ SQL_ID : "cwju3syt2mx9a" PLAN_HASH_VALUE: "31035937"
檢查生成的profile執行文件,需要修改force_match => TRUE ,并且在新的數據庫執行前,還需要針對新的數據庫的testcase用戶名,修改其中對應的表的用戶名。
vi coe_xfr_sql_profile_cwju3syt2mx9a_31035937.sql
~~~~~~~~第三部分,導入性能差的SQLT到11.2.0.2,并查看執行計劃~~~~~~~~~~~~~~
4. 導入SQLT_product到11.2.0.2上,查看11.2.0.2上這個SQL的執行計劃
SQL> SPO OFF; SQL> START tc.sql SQL> REM Executes SQL on TC then produces execution plan. Just execute "@tc.sql" from sqlplus. SQL> SET APPI OFF SERVEROUT OFF; SQL> @@q.sql SQL> SELECT Distinct fc_acc.blnctr_acc, 2 fc_acc.shortcut, 3 fc_acc.acc_name, 4 fc_subject.amount_money_sign, 5 fc_acc.subject_code, 6 fc_acc.corp_code, 7 fc_acc.net_code 8 FROM TC34646.fc_acc, TC34646.fc_user_corp, TC34646.fc_subject 9 WHERE ((fc_acc.corp_code = fc_user_corp.corp_code) OR 10 (acc_cussent is not Null or acc_cussent <> '')) 11 AND (fc_acc.subject_code = fc_subject.subject_code) 12 and ((fc_acc.checker_code is Not Null)) 13 AND (fc_acc.acc_state <> 'R') 14 AND (fc_acc.cur_code = 01) 15 and (fc_acc.acc_state <> 'D') 16 AND (fc_acc.net_code = 1000) 17 AND (fc_acc.acc_type_code = '02' or fc_subject.sys_subject_Code = '1010' or 18 (union_acc is not Null or union_acc <> '')); no rows selected SQL> @@plan.sql SQL> REM Displays plan for most recently executed SQL. Just execute "@plan.sql" from sqlplus. SQL> SET PAGES 2000 LIN 180; SQL> SPO plan.log; SQL> --SELECT * FROM TABLE(DBMS_XPLAN.DISPLAY_CURSOR); SQL> SELECT * FROM TABLE(DBMS_XPLAN.DISPLAY_CURSOR(NULL,NULL,'BASIC ROWS COST PREDICATE')); PLAN_TABLE_OUTPUT ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ EXPLAINED SQL STATEMENT: ------------------------ SELECT Distinct fc_acc.blnctr_acc, fc_acc.shortcut, fc_acc.acc_name, fc_subject.amount_money_sign, fc_acc.subject_code, fc_acc.corp_code, fc_acc.net_code FROM TC34646.fc_acc, TC34646.fc_user_corp, TC34646.fc_subject WHERE ((fc_acc.corp_code = fc_user_corp.corp_code) OR (acc_cussent is not Null or acc_cussent <> '')) AND (fc_acc.subject_code = fc_subject.subject_code) and ((fc_acc.checker_code is Not Null)) AND (fc_acc.acc_state <> 'R') AND (fc_acc.cur_code = 01) and (fc_acc.acc_state <> 'D') AND (fc_acc.net_code = 1000) AND (fc_acc.acc_type_code = '02' or fc_subject.sys_subject_Code = '1010' or (union_acc is not Null or union_acc <> '')) Plan hash value: 561502789 -------------------------------------------------------------------- | Id | Operation | Name | Rows | Cost (%CPU)| -------------------------------------------------------------------- | 0 | SELECT STATEMENT | | | 278K(100)| | 1 | HASH UNIQUE | | 7469K| 278K (1)| |* 2 | HASH JOIN | | 7469K| 3321 (6)| |* 3 | TABLE ACCESS FULL | FC_ACC | 1156 | 38 (3)| | 4 | MERGE JOIN CARTESIAN| | 4189K| 3197 (3)| | 5 | TABLE ACCESS FULL | FC_SUBJECT | 636 | 5 (0)| | 6 | BUFFER SORT | | 6587 | 3192 (3)| | 7 | TABLE ACCESS FULL | FC_USER_CORP | 6587 | 5 (0)| -------------------------------------------------------------------- Predicate Information (identified by operation id): --------------------------------------------------- 2 - access("FC_ACC"."SUBJECT_CODE"="FC_SUBJECT"."SUBJECT_CODE") filter((("FC_ACC"."CORP_CODE"="FC_USER_CORP"."CORP_CODE" OR "ACC_CUSSENT" IS NOT NULL OR "ACC_CUSSENT"<>'') AND ("FC_ACC"."ACC_TYPE_CODE"='02' OR "FC_SUBJECT"."SYS_SUBJECT_CODE"='1010' OR "UNION_ACC" IS NOT NULL OR "UNION_ACC"<>''))) 3 - filter((TO_NUMBER("FC_ACC"."CUR_CODE")=1 AND "FC_ACC"."ACC_STATE"<>'D' AND "FC_ACC"."CHECKER_CODE" IS NOT NULL AND "FC_ACC"."ACC_STATE"<>'R' AND TO_NUMBER("FC_ACC"."NET_CODE")=1000)) 42 rows selected. SQL> SPO OFF;
5. 在將11.2.0.1的profile導入到11.2.0.2,然后查看profile是否生效
執行coe_xfr_sql_profile_cwju3syt2mx9a_31035937.sql前,選需要修改SQL文本為
SELECT Distinct fc_acc.blnctr_acc, fc_acc.shortcut, fc_acc.acc_name, fc_subject.amount_money_sign, fc_acc.subject_code, fc_acc.corp_code, fc_acc.net_code FROM TC34646.fc_acc, <<<<<<<<<<<<<<<<<<<<增加用戶TC34646 TC34646.fc_user_corp, TC34646.fc_subject
6. 執行coe腳本
sqlplus / as sysdba SQL>@/home/oracle/Bright-temp/sqlt_s39917_xecute_TEST/coe_xfr_sql_profile_cwju3syt2mx9a_31035937.sql
7. 再次執行tc.sql
SQL> CONN TC34646/TC34646 SQL> START sqlt_s34646_set_cbo_env.sql SQL> START tc.sql SQL> REM Executes SQL on TC then produces execution plan. Just execute "@tc.sql" from sqlplus. SQL> SET APPI OFF SERVEROUT OFF; SQL> @@q.sql SQL> SELECT Distinct fc_acc.blnctr_acc, 2 fc_acc.shortcut, 3 fc_acc.acc_name, 4 fc_subject.amount_money_sign, 5 fc_acc.subject_code, 6 fc_acc.corp_code, 7 fc_acc.net_code 8 FROM TC34646.fc_acc, TC34646.fc_user_corp, TC34646.fc_subject 9 WHERE ((fc_acc.corp_code = fc_user_corp.corp_code) OR 10 (acc_cussent is not Null or acc_cussent <> '')) 11 AND (fc_acc.subject_code = fc_subject.subject_code) 12 and ((fc_acc.checker_code is Not Null)) 13 AND (fc_acc.acc_state <> 'R') 14 AND (fc_acc.cur_code = 01) 15 and (fc_acc.acc_state <> 'D') 16 AND (fc_acc.net_code = 1000) 17 AND (fc_acc.acc_type_code = '02' or fc_subject.sys_subject_Code = '1010' or 18 (union_acc is not Null or union_acc <> '')); no rows selected SQL> @@plan.sql SQL> REM Displays plan for most recently executed SQL. Just execute "@plan.sql" from sqlplus. SQL> SET PAGES 2000 LIN 180; SQL> SPO plan.log; SQL> --SELECT * FROM TABLE(DBMS_XPLAN.DISPLAY_CURSOR); SQL> SELECT * FROM TABLE(DBMS_XPLAN.DISPLAY_CURSOR(NULL,NULL,'BASIC ROWS COST PREDICATE')); PLAN_TABLE_OUTPUT ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ EXPLAINED SQL STATEMENT: ------------------------ SELECT Distinct fc_acc.blnctr_acc, fc_acc.shortcut, fc_acc.acc_name, fc_subject.amount_money_sign, fc_acc.subject_code, fc_acc.corp_code, fc_acc.net_code FROM TC34646.fc_acc, TC34646.fc_user_corp, TC34646.fc_subject WHERE ((fc_acc.corp_code = fc_user_corp.corp_code) OR (acc_cussent is not Null or acc_cussent <> '')) AND (fc_acc.subject_code = fc_subject.subject_code) and ((fc_acc.checker_code is Not Null)) AND (fc_acc.acc_state <> 'R') AND (fc_acc.cur_code = 01) and (fc_acc.acc_state <> 'D') AND (fc_acc.net_code = 1000) AND (fc_acc.acc_type_code = '02' or fc_subject.sys_subject_Code = '1010' or (union_acc is not Null or union_acc <> '')) Plan hash value: 31035937 ------------------------------------------------------------------------------ | Id | Operation | Name | Rows | Cost (%CPU)| ------------------------------------------------------------------------------ | 0 | SELECT STATEMENT | | | 8755 (100)| | 1 | HASH UNIQUE | | 7469K| 8755 (1)| | 2 | CONCATENATION | | | | | 3 | MERGE JOIN CARTESIAN | | 226K| 245 (3)| | 4 | NESTED LOOPS | | | | | 5 | NESTED LOOPS | | 34 | 73 (2)| |* 6 | TABLE ACCESS FULL | FC_ACC | 35 | 38 (3)| |* 7 | INDEX UNIQUE SCAN | PK_FC_SUBJECT | 1 | 0 (0)| |* 8 | TABLE ACCESS BY INDEX ROWID| FC_SUBJECT | 1 | 1 (0)| | 9 | BUFFER SORT | | 6587 | 244 (3)| | 10 | TABLE ACCESS FULL | FC_USER_CORP | 6587 | 5 (0)| | 11 | MERGE JOIN CARTESIAN | | 2285 | 46 (3)| | 12 | NESTED LOOPS | | | | | 13 | NESTED LOOPS | | 1 | 39 (3)| |* 14 | TABLE ACCESS FULL | FC_ACC | 1 | 38 (3)| |* 15 | INDEX UNIQUE SCAN | PK_FC_SUBJECT | 1 | 0 (0)| |* 16 | TABLE ACCESS BY INDEX ROWID| FC_SUBJECT | 1 | 1 (0)| | 17 | BUFFER SORT | | 6587 | 45 (3)| | 18 | TABLE ACCESS FULL | FC_USER_CORP | 6587 | 7 (0)| |* 19 | HASH JOIN | | 193 | 51 (4)| |* 20 | HASH JOIN | | 11 | 43 (3)| |* 21 | TABLE ACCESS FULL | FC_ACC | 11 | 38 (3)| | 22 | TABLE ACCESS FULL | FC_SUBJECT | 636 | 5 (0)| | 23 | TABLE ACCESS FULL | FC_USER_CORP | 6587 | 7 (0)| ------------------------------------------------------------------------------ Predicate Information (identified by operation id): --------------------------------------------------- 6 - filter((TO_NUMBER("FC_ACC"."CUR_CODE")=1 AND "ACC_CUSSENT"<>'' AND "FC_ACC"."ACC_STATE"<>'D' AND "FC_ACC"."CHECKER_CODE" IS NOT NULL AND "FC_ACC"."ACC_STATE"<>'R' AND TO_NUMBER("FC_ACC"."NET_CODE")=1000)) 7 - access("FC_ACC"."SUBJECT_CODE"="FC_SUBJECT"."SUBJECT_CODE") 8 - filter(("FC_ACC"."ACC_TYPE_CODE"='02' OR "FC_SUBJECT"."SYS_SUBJECT_CODE"='1010' OR "UNION_ACC" IS NOT NULL OR "UNION_ACC"<>'')) 14 - filter(("ACC_CUSSENT" IS NOT NULL AND TO_NUMBER("FC_ACC"."CUR_CODE")=1 AND "FC_ACC"."ACC_STATE"<>'D' AND "FC_ACC"."CHECKER_CODE" IS NOT NULL AND "FC_ACC"."ACC_STATE"<>'R' AND TO_NUMBER("FC_ACC"."NET_CODE")=1000 AND LNNVL("ACC_CUSSENT"<>''))) 15 - access("FC_ACC"."SUBJECT_CODE"="FC_SUBJECT"."SUBJECT_CODE") 16 - filter(("FC_ACC"."ACC_TYPE_CODE"='02' OR "FC_SUBJECT"."SYS_SUBJECT_CODE"='1010' OR "UNION_ACC" IS NOT NULL OR "UNION_ACC"<>'')) 19 - access("FC_ACC"."CORP_CODE"="FC_USER_CORP"."CORP_CODE") 20 - access("FC_ACC"."SUBJECT_CODE"="FC_SUBJECT"."SUBJECT_CODE") filter(("FC_ACC"."ACC_TYPE_CODE"='02' OR "FC_SUBJECT"."SYS_SUBJECT_CODE"='1010' OR "UNION_ACC" IS NOT NULL OR "UNION_ACC"<>'')) 21 - filter((TO_NUMBER("FC_ACC"."CUR_CODE")=1 AND "FC_ACC"."ACC_STATE"<>'D' AND "FC_ACC"."CHECKER_CODE" IS NOT NULL AND "FC_ACC"."ACC_STATE"<>'R' AND TO_NUMBER("FC_ACC"."NET_CODE")=1000 AND LNNVL("ACC_CUSSENT" IS NOT NULL) AND LNNVL("ACC_CUSSENT"<>''))) 74 rows selected. SQL> SPO OFF; SQL>
最終,可以看到,在11.2.0.2上,也按照11.2.0.1上的執行計劃進行執行SQL了。
這里要強調一下SQLT真的是非常使用的SQL 調優工具。而且非常易用,很多步驟都在SQLT的HTML文件中,逐步指導。建議大家學習使用。
參考文檔:
FAQ: SQLT (SQLTXPLAIN) Frequently Asked Questions (Doc ID 1454160.1)
All About the SQLT Diagnostic Tool (Doc ID 215187.1)
How to Collect Standard Diagnostic Information Using SQLT for SQL Issues (Doc ID 1683772.1)
SQLT Usage Instructions (Doc ID 1614107.1)
How to Generate a SQLT of a SQL Statement That Hangs or Raises an Error (Doc ID 2146816.1)
How to Use SQLT (SQLTXPLAIN) to Create a Testcase Without Row Data (Doc ID 1470811.1)
免責聲明:本站發布的內容(圖片、視頻和文字)以原創、轉載和分享為主,文章觀點不代表本網站立場,如果涉及侵權請聯系站長郵箱:is@yisu.com進行舉報,并提供相關證據,一經查實,將立刻刪除涉嫌侵權內容。