DB2 - Problem description
Problem IT02048 | Status: Closed |
INACCURATE STATS COLLECTED BY AUTORUNSTATS WHEN IMPORT COMMAND RUNS FOR MORE THAN 2 HOURS, HOLDING THE LOCKS ON CATALOG TABLES | |
product: | |
DB2 FOR LUW / DB2FORLUW / A50 - DB2 | |
Problem description: | |
Autorunstats report far less rows than reality. This issue is reproduced when both auto_runstats and auto_stmt_stats db cfg parameter is set to ON To reproduce 1. create table s1.tab1 (a int, b int) 2 import from data.ixf of ixf commitcount 10000 insert into s1.tab1 3. import finished. over 5 million rows inserted. Import command took 4 hours to complete. 4. card is reported as only 1 million rows, instead of 5 million. Statistics are not updated even after waiting for one whole day. db2 => select card from syscat.tables where tabname='TAB1' CARD -------------------- 1041440 1 record(s) selected. db2 => select count(*) from s1.tab1 1 ----------- 5905110 1 record(s) selected. The following error message can be observed in db2dump/events/db2optstats.0.log file, repeatedly for every 5 min of time interval in case the above scenario is hit. 2014-06-03-16.46.05.454131-240 E181114E709 LEVEL: Event PID : 20235 TID : 46913147627840 KTID : 20823 PROC : db2sysc 0 INSTANCE: sahanant NODE : 000 DB : D10 APPHDL : 0-259 APPID: *N0.sahanant.140603204609 AUTHID : SAHANANT HOSTNAME: hotel61 EDUID : 363 EDUNAME: db2agent (D10) 0 FUNCTION: DB2 UDB, relation data serv, sqlr_upd_cat_except_systables, probe:22463 WRITE : TABLE AND INDEX STATS : Object name with schema : AT "2014-06-03-16.46.05.453968" : BY "Asynchronous" : start OBJECT : Object name with schema, 22 bytes S1 .TAB1 IMPACT : None DATA #1 : String, 10 bytes RTS Daemon 2014-06-03-16.46.10.479974-240 I181824E566 LEVEL: Error PID : 20235 TID : 46913147627840 KTID : 20823 PROC : db2sysc 0 INSTANCE: sahanant NODE : 000 DB : D10 APPHDL : 0-259 APPID: *N0.sahanant.140603204609 AUTHID : SAHANANT HOSTNAME: hotel61 EDUID : 363 EDUNAME: db2agent (D10) 0 FUNCTION: DB2 UDB, relation data serv, sqlr_update_catalogs, probe:26964 MESSAGE : ZRC=0x80100044=-2146435004=SQLP_LTIMEOUT "LockTimeOut - tran rollback Reason code 68" 2014-06-03-16.46.10.494621-240 E182391E814 LEVEL: Event PID : 20235 TID : 46913147627840 KTID : 20823 PROC : db2sysc 0 INSTANCE: sahanant NODE : 000 DB : D10 APPHDL : 0-259 APPID: *N0.sahanant.140603204609 AUTHID : SAHANANT HOSTNAME: hotel61 EDUID : 363 EDUNAME: db2agent (D10) 0 FUNCTION: DB2 UDB, relation data serv, sqlr_update_catalogs, probe:27425 WRITE : TABLE AND INDEX STATS : Object name with schema : AT "2014-06-03-16.46.10.494466" : BY "Asynchronous" : failure OBJECT : Object name with schema, 22 bytes S1 .TAB1 IMPACT : None DATA #1 : String, 10 bytes | |
Problem Summary: | |
**************************************************************** * USERS AFFECTED: * * software * **************************************************************** * PROBLEM DESCRIPTION: * * statistics collected by automatic staitstics is not * * up-to-date * **************************************************************** * RECOMMENDATION: * * upgrade to v10.5 fp4 * **************************************************************** | |
Local Fix: | |
re-run the runstats command manually after the IMPORT command completed | |
available fix packs: | |
DB2 Cancun Release 10.5.0.4 (also known as Fix Pack 4) for Linux, UNIX, and Windows | |
Solution | |
fixed in version v10.5 fp4 | |
Workaround | |
not known / see Local fix | |
Timestamps | |
Date - problem reported : Date - problem closed : Date - last modified : | 27.05.2014 08.09.2014 08.09.2014 |
Problem solved at the following versions (IBM BugInfos) | |
Problem solved according to the fixlist(s) of the following version(s) | |
10.5.0.4 |