E1: KER: PKG: Hung Call Object Kernel In DeadLock State Blocks Package Deployment Due To Object Tracking
(Doc ID 2451638.1)
Last updated on FEBRUARY 24, 2020
Applies to:JD Edwards EnterpriseOne Tools - Version 9.2 and later
Information in this document applies to any platform.
Call Object kernel on Enterprise server is going into zombie status, and when this happens the client is unable to deploy any packages.
Development has identified that this is caused when "Object Usage Tracking" is enabled, the OUT thread deadlocks with the free environment thread.
Example from Enterprise server CallStack:
Call stack from hung kernel shows thread "386", and, "389" causing a deadlock.
1) Thread 386 is processing free environment request (jdbFreeEnvInternal()) and has acquired JDB_INITFREEENV mutex.
2) Simultaneously , thread 389 is kicked loose ¿to execute ¿jdeInitiateCallObjectOUTHashClear()¿ , it acquires JDBLOCK_OUTUSER , and waits for JDB_INITFREEENV ( to initialize static environment).
3) Meanwhile thread 386 ¿ while executing ¿jdbFreeEnvInternal()¿ tries to get JDBLOCK_OUTUSER which is now held by thread 389.
4) In short , thread 386 cannot proceed as long as 389 is holding JDBLOCK_OUTUSER , and 389 is hung on JDB_INITFREEENV ¿ which is held by 386.
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