gc cr block lost

故障现象
1月3日上午10时，一客户数据库实例1重启，当业务切换到实例2时，实例2也重启。
故障分析
日志分析：

下面信息摘取自LMON trace

*** 2013-01-0310:04:12.203

kjfmrcvrchk:receiver LMS[4] has no heartbeat for 251 sec (1357178400.1357178651.0).

kjfmrcvrchk:receiver LMS[4] not in running mode

kjfmrcvrchk:Dumping callstack of lms4

Submittingasynchronized dump request [20]

kjfmrcvrchk:receivers are not healthy. kill instance.

ksuitm: waitingup to [5] seconds before killing DIAG(13789)

从以上LMON TRACE中可以看出10:04:12检测到进程LMS失去心跳251秒，5秒后将kill实例。因此从实例1的告警日志中可以看出，数据库在10:04:17时报LMON detects unhealthy receivers,被LMON进程kill的信息，详细信息如下：

Thu Jan 0309:55:11 EAT 2013

Thread 1advanced to log sequence 9754 (LGWR switch)

  Current log# 1 seq# 9754 mem# 0:/vghn03/oradata/esshn/vghn03_1_rd12.log

  Current log# 1 seq# 9754 mem# 1:/vghn02/oradata/esshn/vghn02_1_rd11.log

Thu Jan 0310:04:17 EAT 2013

LMON detectsunhealthy receivers.

Please checkLMON and DIAG trace files for detail.

Thu Jan 0310:04:17 EAT 2013

LMON (ospid:13793) is terminating the instance.

LMON:terminating instance due to error 481