Informix - Problem description
Problem IT20075 | Status: Closed |
RSS PRIMARY SERVER CAN HANG/LOCK UP WITH RSS_SEND/RSS_RECV THREADS SPINNING ON THE BACKLOGWAITLOCK SPIN LOCK | |
product: | |
INFORMIX SERVER / 5725A3900 / C10 - IDS 12.10 | |
Problem description: | |
This is the same problem as APAR IC78772. The primary can hang running constantly on the cpu vps trying to acquire a spin lock. Both RSS_Send and RSS_recv threads can hang. The RSS_Send will wait on spin lock in getPagesFromCache() and RSS_Recv in cloneSourceRecvThread(). The thread holding the backlogWaitLock spin lock will have a stack like: yield_processor_mvp() mt_lock_wait() mt_lock() cloneCacheLogPageInt() cloneCacheInt() cloneCacheLogPage() logwrite() log_put() logm_write() logputx() rscommit() iscommit() sqiscommit() sqcommit() aud_sqcommit() sql_commit() sq_commit() sqmain() listen_verify() spawn_thread() th_init_initgls() startup() The defect occurs with 1 or 2 CPU VPs. Since the threads are running constantly in the CPU VPs, the sqlexec thread cannot continue even if the mutex is already free. | |
Problem Summary: | |
**************************************************************** * USERS AFFECTED: * * All users * **************************************************************** * PROBLEM DESCRIPTION: * * See Error Description * **************************************************************** * RECOMMENDATION: * * Update to IBM Informix Server 12.10.xC9 * **************************************************************** | |
Local Fix: | |
Configure 3 or more CPU VPs. | |
Solution | |
Workaround | |
not known / see Local fix | |
Timestamps | |
Date - problem reported : Date - problem closed : Date - last modified : | 05.04.2017 10.07.2017 10.07.2017 |
Problem solved at the following versions (IBM BugInfos) | |
12.10.xC9 | |
Problem solved according to the fixlist(s) of the following version(s) |