Tuxedo Backup Node In MP Configuration Hangs When Primary Node Crashes (Doc ID 1224825.1)

Last updated on MAY 04, 2023

Applies to:

Symptoms

MP Tuxedo configuration :
MASTER on AIX with TUXEDO 10.3 with RP22
SLAVE on Linux x86_64 with TUXEDO 10.3 with RP 22

- GWTDOMAIN on both machines connected to 2 other domains that receive client requests.
- After starting a load test, around 50 clients connected to each frontend machine that do 1000 calls to services published on MP domain.
- Shutdown the machine where master is running to simulate the machine crash (shutdown -K)
- After few minutes all clients receives TPETIME errors.
- strace -o BBL.out -p PID_of_BBL shows that the  BBL is locked on : msgrcv
- "tmadmin -r" on the backup machine does not reply.
- All tuxedo servers are hanging trying to get a lock on Tuxedo Bulletin Board

Changes

Cause

My Oracle Support provides customers with access to over a million knowledge articles and a vibrant support community of peers and Oracle experts.