ASE: local Director Error: connect to Director on localhost failed

[郭尚中_1]

Cluster: 2 GS60E+RA800
OS: 4.0F
ASE: 1.6
ORACLE8i: OPS

When 1 system crash for hardware error. The other system can not be connected use oracle client program. Only restart oracle listener can fix this problem.
This problem will not occur when 1 system shutdown normally.

The following are daemon logs from the good system when the other machines crash.

Mar 28 21:50:36 gs60e01 ASE: mcgs60e01 Agent Notice: agent on mcgs60e02 should start director, but isn't in RUN state
Mar 28 21:50:36 gs60e01 ASE: mcgs60e01 Agent Notice: starting a new director
Mar 28 21:50:36 gs60e01 ASE: mcgs60e01 Agent Notice: client process hung up during connect phase
Mar 28 21:50:37 gs60e01 ASE: mcgs60e01 AseMgr Error: Director hasn't started up yet or local network is disconnected ...
Mar 28 21:50:52 gs60e01 last message repeated 3 times
Mar 28 21:50:57 gs60e01 ASE: local Director Error: timeout connecting to Director on localhost
Mar 28 21:50:57 gs60e01 ASE: local Director Error: connect to Director on localhost failed
Mar 28 21:50:57 gs60e01 ASE: local Director Notice: acceptConnect: not accepting connections!
Mar 28 21:50:58 gs60e01 last message repeated 9052 times
Mar 28 21:50:58 gs60e01 ASE: local Director ***ALERT: Member mcgs60e02 is not available
Mar 28 21:50:58 gs60e01 ASE: local Director Notice: acceptConnect: not accepting connections!
Mar 28 21:51:05 gs60e01 last message repeated 44957 times
Mar 28 21:51:05 gs60e01 ASE: mcgs60e01 AseMgr Notice: reconnected to director
Mar 28 21:51:05 gs60e01 ASE: mcgs60e01 Director Notice: client process hung up during connect phase

Seems asedirector can not be started when the other system crash.

Does anyone has solution for this situation?

Thanks!

Guo ShangZhong