Concurrent Manager Crashes FNDLIBR or INVLIBR Core Dumps Continuously and Get Terminated
(Doc ID 1506643.1)
Last updated on AUGUST 13, 2020
Applies to:Oracle Concurrent Processing - Version 220.127.116.11 to 12.1.3 [Release 11.5 to 12.1]
Information in this document applies to any platform.
The Concurrent Manager (FNDLIBR) crashes regularly with error.
Manager process: spid=(6540:2672), cpid=(56660), ORA pid=(56) manager=(0/10277)
Received lock, set mgralive=N
Manager process NOT reconnecting, mgralive=(N)
Manager was running a request, id=1974072
CONNECTING_PROCESS -- Could not mark PROCESS_STATUS_CODE from A to C - cpid (56660)
Back from connecting_process(), a_to_c=(FALSE).
Convinced now that manager is dead. About to actually cleanup request.
Found running request 1974072 attached to dead manager process.
Setting request status to completed.
In some cases Inventory Manager (INVMGR) INVLIBR is getting terminated only on Clone instance when Production instance is working fine.
20-JAN-2014 03:19:12 - FDPRRC- Query node, instance from parent request for resubmission
20-JAN-2014 03:19:12 - FDPRRC- node=NULL, instance=NULL from parent request for resubmission
20-JAN-2014 03:19:12 - FDPRRC- Query node, instance for program INCTCM
20-JAN-2014 03:19:12 - FDPRRC- node=NULL, instance=NULL for program INCTCM
sh: line 68: 30008 Segmentation fault FNDLIBR FND CPMGR "FNDCPMBR sysmgr=\"\" sleep=30 pmon=4 quesiz=1 diag=y logfile=/XX/XX/XX/logs/appl/conc/log/ebs01_0120.mgr " <
After setting ulimit -a unlimited, It generates core file under $APPLCSF/$APPLLOG . When stack trace is taken , using OS tool (gdb), It was found that the symbol fdugpi in the stack trace.
Limited number of customers applying any patch containing fdugpi version as 115.7 (in 11i) 120.1.12000000.4 ( in 12.0.x) and 120.3.12010000.2 (in 12.1.x) can cause concurrent managers to core dump on some systems.
To view full details, sign in with your My Oracle Support account.
Don't have a My Oracle Support account? Click to get started!
In this Document