DB2 - Problem description
Problem IC73185 | Status: Closed |
DB2 MIGHT ENCOUNTER 'DEAD LATCH' ISSUE AND INSTANCE WILL HANG. | |
product: | |
DB2 FOR LUW / DB2FORLUW / 950 - DB2 | |
Problem description: | |
To identify the problem, issue 'db2fodc -hang' command to generate diagnostic information. Then in stack files you will see 2 agents similar to the followings: ================================ <StackTrace> sqloXlatchConflict sqlpm_write_appl_locks sqm_snap_appl_locks sqm_snap_db_locks sqlmonssagnt <LatchInformation> Waiting on latch type: (SQLO_LT_SQLP_TENTRY__tranEntryLatch) - Address: (7000001bc956e68), Line: 748, File: /view/db2_v97fp2_aix64_s100514/vbs/engn/include/sqlpi_inlines.h Holding Latch type: (SQLO_LT_sqeLocalDatabase__dblatch) - Address: (780000000780328), Line: 395, File: sqlmutil.C HoldCount: 1 Holding Latch type: (SQLO_LT_sqeAppServices__m_appServLatch) - Address: (7800000002f04f8), Line: 7822, File: sqlmonss.C HoldCount: 1 Holding Latch type: (SQLO_LT_sqeApplication__masterAppLatch) - Address: (7000002513101c4), Line: 327, File: /view/db2_v97fp2_aix64_s100514/vbs/engn/include/sqle_app_service s.h HoldCount: 1 ... <StackTrace> sqloXlatchConflict sqlkdDispatchRequest sqlkdInterrupt sqlrr2pd sqlrkcmt sqlrkcm sqlrrcom_dps sqlrr_xcommit <LatchInformation> Waiting on latch type: (SQLO_LT_sqeApplication__masterAppLatch) - Address: (7000002513101c4), Line: 997, File: /view/db2_v97fp2_aix64_s100514/vbs/engn/include/sqle_application .h Holding Latch type: (SQLO_LT_SQLP_TENTRY__tranEntryLatch) - Address: (7000001bc956e68), Line: 748, File: /view/db2_v97fp2_aix64_s100514/vbs/engn/include/sqlpi_inlines.h HoldCount: 1 ================================ The deadlatch is constructed by 'masterAppLatch' and 'tranEntryLatch'. You will have to recycle the instance to get out of the 'dead latch' situation. | |
Problem Summary: | |
**************************************************************** * USERS AFFECTED: * * ALL * **************************************************************** * PROBLEM DESCRIPTION: * * To identify the problem, issue 'db2fodc -hang' command to * * generate diagnostic information. Then in stack files you * * will see 2 agents similar to the followings: * * ================================ * * <StackTrace> * * sqloXlatchConflict * * sqlpm_write_appl_locks * * sqm_snap_appl_locks * * sqm_snap_db_locks * * sqlmonssagnt * * * * <LatchInformation> * * Waiting on latch type: (SQLO_LT_SQLP_TENTRY__tranEntryLatch) * * - Address: (7000001bc956e68), Line: 748, File: * * /view/db2_v97fp2_aix64_s100514/vbs/engn/include/sqlpi_inlines. * * * Holding Latch type: (SQLO_LT_sqeLocalDatabase__dblatch) - * * Address: (780000000780328), Line: 395, File: sqlmutil.C * * HoldCount: 1 * * Holding Latch type: (SQLO_LT_sqeAppServices__m_appServLatch) * * - Address: (7800000002f04f8), Line: 7822, File: sqlmonss.C * * HoldCount: 1 * * Holding Latch type: (SQLO_LT_sqeApplication__masterAppLatch) * * - Address: (7000002513101c4), Line: 327, File: * * /view/db2_v97fp2_aix64_s100514/vbs/engn/include/sqle_app_servi * HoldCount: 1 * * * * ... * * * * <StackTrace> * * sqloXlatchConflict * * sqlkdDispatchRequest * * sqlkdInterrupt * * sqlrr2pd * * sqlrkcmt * * sqlrkcm * * sqlrrcom_dps * * sqlrr_xcommit * * * * <LatchInformation> * * Waiting on latch type: * * (SQLO_LT_sqeApplication__masterAppLatch) - Address: * * (7000002513101c4), Line: 997, File: * * /view/db2_v97fp2_aix64_s100514/vbs/engn/include/sqle_applicati * * * Holding Latch type: (SQLO_LT_SQLP_TENTRY__tranEntryLatch) - * * Address: (7000001bc956e68), Line: 748, File: * * /view/db2_v97fp2_aix64_s100514/vbs/engn/include/sqlpi_inlines. * HoldCount: 1 * * ================================ * * The deadlatch is constructed by 'masterAppLatch' and * * 'tranEntryLatch'. * * * * You will have to recycle the instance to get out of the * * 'dead latch' situation. * **************************************************************** * RECOMMENDATION: * * Upgrade to db2 version 9.7 fp 4 * **************************************************************** | |
Local Fix: | |
The problem only occurs when you're getting snapshot. So it can be avoided if you don't get any snapshot. | |
Solution | |
Workaround | |
not known / see Local fix | |
Timestamps | |
Date - problem reported : Date - problem closed : Date - last modified : | 10.12.2010 10.12.2010 10.12.2010 |
Problem solved at the following versions (IBM BugInfos) | |
9.7.FP4 | |
Problem solved according to the fixlist(s) of the following version(s) |