informix经常down，各位看看是什么原因？

boyfsl

环境：informix版本11.50.fc3，双cpu的服务器
报错如下：
09:28:41  Assert Failed: Exception Caught. Type: MT_EX_OS, Context: mem
09:28:41  IBM Informix Dynamic Server Version 11.50.FC3
09:28:41   Who: Session(587712, shopuser@192.168.1.7, 2368, 0x5b78701
                Thread(590710, sqlexec, 5b757fe8, 1)
                File: mtex.c Line: 412
09:28:41   Action: Please notify IBM Informix Technical Support.
09:28:41  stack trace for pid 2602 written to /home/informix/tmp/af.75e63c9
09:28:41   See Also: /home/informix/tmp/af.75e63c9, shmem.75e63c9.0
09:28:43  Recursive Exception - Server exiting
09:28:43  invoke_alarm(): /bin/sh -c '/home/informix/etc/alarmprogram.sh 5 6 "Internal Subsystem failure: 'MT'" "Recursive Exception - Server exiting" ""'
09:28:43  invoke_alarm(): mt_exec failed, status 32256, errno 0
09:28:44  The Master Daemon Died
09:28:44  invoke_alarm(): /bin/sh -c '/home/informix/etc/alarmprogram.sh 5 6 "Internal Subsystem failure: 'MT'" "The Master Daemon Died" ""'
09:28:44  invoke_alarm(): mt_exec failed, status 32256, errno 0
09:28:44  PANIC: Attempting to bring system down
09:33:25  IBM Informix Dynamic Server Started.
09:33:25  The number of configured CPU poll threads exceeds 'NUMCPUVPS'.
          NETTYPE 'onlf' poll threads started on NET VPs.

找了半天找不到什么原因，各位帮忙分析下是哪里的问题？

boyfsl

我能不能把ALARMPROGRAM $INFORMIXDIR/etc/alarmprogram.sh给屏蔽了？
是不是执行了alarmprogram.sh，把数据库给down了？

liaosnet

两个建议：
1，升级到11.50.FC9
2，NETTYPE 的参数中的线程数量减少到VPCLASS cpu中指定的数量。。

大梦

简单的说就是Informix BUG
要么升级要么看看af文件里面的sql语句,能不能从应用层面避免.