RAC hang up after interface down

Hello All,

I have following problem, may be any guru knows whats happening. Env:
Solaris 9
Veritas 4.1
Oracle 9.2.0.7

After interface down all nodes (I have 2 nodes) all nodes hang up.

 load averages: 4.97, 5.05, 4.89
19:04:02
89 processes: 83 sleeping, 4 running, 2 on cpu CPU states: 0.0% idle, 98.6% user, 1.4% kernel, 0.0% iowait, 0.0% swap
Memory: 2048M real, 423M free, 1271M swap in use, 4251M swap free

   PID USERNAME LWP PRI NICE SIZE RES STATE TIME CPU COMMAND

  7988 oracle     1  30    0 1108M 1069M cpu/0   24:55 20.80% oracle
  8003 oracle     1  31    0 1108M 1063M run     24:49 20.25% oracle
  7968 oracle     1  22    0 1109M 1064M run     24:30 20.14% oracle
  7978 oracle     1  31    0 1111M 1067M run     17:36 18.86% oracle
  7990 oracle     1  31    0 1108M 1068M run     17:28 18.79% oracle



oracle_at_rdb1> ps -ef | egrep "7988|8003|7968|7978|7990"
  oracle  7988     1 20   Jul 04 ?       25:15 ora_smon_OMP1
  oracle  7990     1 20   Jul 04 ?       17:49 ora_reco_OMP1
  oracle  7968     1 19   Jul 04 ?       24:49 ora_lmon_OMP1
  oracle  8003     1 21   Jul 04 ?       25:10 ora_lck0_OMP1
  oracle  7978     1 20   Jul 04 ?       17:56 ora_lms0_OMP1

cat alert.log

Wed Jul 5 18:08:39 2006
IPC Send timeout detected. Sender ospid 8003 Wed Jul 5 18:08:48 2006
IPC Send timeout detected. Sender ospid 7988 Wed Jul 5 18:09:06 2006
IPC Send timeout detected. Sender ospid 7968 Wed Jul 5 18:19:37 2006
IPC Send timeout detected. Sender ospid 7978 Wed Jul 5 18:19:48 2006
IPC Send timeout detected. Sender ospid 7990 Wed Jul 5 18:57:42 2006
>>> WAITED TOO LONG FOR A ROW CACHE ENQUEUE LOCK! pid=14

But interface is already UP.

As I understand IPC does not recognise it.

Almost the same on another node.

HELP :) Received on Wed Jul 05 2006 - 10:08:46 CDT