oracle9i rac for aix,AIX5.3+ORACLE9IRAC 的安装

OS：AIX5300-02

HA：HACMP5.2

ORACLE：ORA9201

在CREATE DATABASE的时候，一个节点的GSD可以起来，另一个却死活起不来，

报PRKC-1021错误：

PRKC-1021 : Problem in the clusterware

Failed to get list of active nodes from clusterware

下面是我在metalink上的帖子。

on node1 (good node)

$lsnodes -n -v

NODE1

NODE2

# odmget HACMPnode

HACMPnode:

name = "LGJMPTA"

object = "VERBOSE_LOGGING"

value = "high"

node_id = 1

node_handle = 1

version = 7

HACMPnode:

name = "LGJMPTA"

object = "NAME_SERVER"

value = "FALSE"

node_id = 1

node_handle = 1

version = 7

HACMPnode:

name = "LGJMPTA"

object = "DEBUG_LEVEL"

value = "Standard"

node_id = 1

node_handle = 1

version = 7

HACMPnode:

name = "LGJMPTA"

object = "SYS_ADMIN_ID"

value = ""

node_id = 1

node_handle = 1

version = 7

HACMPnode:

name = "LGJMPTA"

object = "MODEM_NUMBER"

value = ""

node_id = 1

node_handle = 1

version = 7

HACMPnode:

name = "LGJMPTA"

object = "MAX_BLOCK_SIZE"

value = "8192"

node_id = 1

node_handle = 1

version = 7

HACMPnode:

name = "LGJMPTA"

object = "COMMUNICATION_PATH"

value = "192.168.10.1"

node_id = 1

node_handle = 1

version = 7

HACMPnode:

name = "LGJMPTB"

object = "VERBOSE_LOGGING"

value = "high"

node_id = 2

node_handle = 2

version = 7

HACMPnode:

name = "LGJMPTB"

object = "MODEM_NUMBER"

value = ""

node_id = 2

node_handle = 2

version = 7

HACMPnode:

name = "LGJMPTB"

object = "MAX_BLOCK_SIZE"

value = "8192"

node_id = 2

node_handle = 2

version = 7

HACMPnode:

name = "LGJMPTB"

object = "COMMUNICATION_PATH"

value = "192.168.10.2"

node_id = 2

node_handle = 2

version = 7

HACMPnode:

name = "LGJMPTB"

object = "NAME_SERVER"

value = "FALSE"

node_id = 2

node_handle = 2

version = 7

HACMPnode:

name = "LGJMPTB"

object = "DEBUG_LEVEL"

value = "Standard"

node_id = 2

node_handle = 2

version = 7

HACMPnode:

name = "LGJMPTB"

object = "SYS_ADMIN_ID"

value = ""

node_id = 2

node_handle = 2

version = 7

on node2 (bad node)

$lsnodes -n -v

lsnodes: cannot initialize cluster interface

skgxn error number 1

operation sskgxn_nm_create

location skgxncin

#odmget HACMPnode

HACMPnode:

name = "LGJMPTA"

object = "VERBOSE_LOGGING"

value = "high"

node_id = 1

node_handle = 1

version = 7

HACMPnode:

name = "LGJMPTA"

object = "NAME_SERVER"

value = "FALSE"

node_id = 1

node_handle = 1

version = 7

HACMPnode:

name = "LGJMPTA"

object = "DEBUG_LEVEL"

value = "Standard"

node_id = 1

node_handle = 1

version = 7

HACMPnode:

name = "LGJMPTA"

object = "SYS_ADMIN_ID"

value = ""

node_id = 1

node_handle = 1

version = 7

HACMPnode:

name = "LGJMPTA"

object = "MODEM_NUMBER"

value = ""

node_id = 1

node_handle = 1

version = 7

HACMPnode:

name = "LGJMPTA"

object = "MAX_BLOCK_SIZE"

value = "8192"

node_id = 1

Text continued in next action...

17-OCT-05 07:43:03 GMT

Text continued from previous action...

node_handle = 1

version = 7

HACMPnode:

name = "LGJMPTA"

object = "COMMUNICATION_PATH"

value = "192.168.10.1"

node_id = 1

node_handle = 1

version = 7

HACMPnode:

name = "LGJMPTB"

object = "VERBOSE_LOGGING"

value = "high"

node_id = 2

node_handle = 2

version = 7

HACMPnode:

name = "LGJMPTB"

object = "MODEM_NUMBER"

value = ""

node_id = 2

node_handle = 2

version = 7

HACMPnode:

name = "LGJMPTB"

object = "MAX_BLOCK_SIZE"

value = "8192"

node_id = 2

node_handle = 2

version = 7

HACMPnode:

name = "LGJMPTB"

object = "COMMUNICATION_PATH"

value = "192.168.10.2"

node_id = 2

node_handle = 2

version = 7

HACMPnode:

name = "LGJMPTB"

object = "NAME_SERVER"

value = "FALSE"

node_id = 2

node_handle = 2

version = 7

HACMPnode:

name = "LGJMPTB"

object = "DEBUG_LEVEL"

value = "Standard"

node_id = 2

node_handle = 2

version = 7

HACMPnode:

name = "LGJMPTB"

object = "SYS_ADMIN_ID"

value = ""

node_id = 2

node_handle = 2

version = 7

So, How can i do?

17-OCT-05 08:06:05 GMT

New info : AHDXJF : I trun on gsdctl tracing on bad node LGJMPTB:

$ gsdctl start

[main] [16:2:27:794] [GSDCTLDriver.startDaemon:153] Going into GetActiveNodes

constructor...

[main] [16:2:27:801] [NativeSystem.:123] Detected Cluster

[main] [16:2:27:803] [NativeSystem.:124] Cluster existence = true

[main] [16:2:27:806] [UnixSystem.:118] Going to load SRVM library

[main] [16:2:27:811] [UnixSystem.:118] loaded libraries

[main] [16:2:27:815] [GetActiveNodes.access$0:256] Inside

GetActiveNodes.initializeCluster

[main] [16:2:27:819] [ClusterResult.:44] NativeResult: The String

obtained is0| skgxncin call failedCould not initialize

[main] [16:2:27:821] [ClusterResult.:44] The status string is: 0

[main] [16:2:27:824] [ClusterResult.:44] The result string is: skgxncin

call failedCould not initialize 1

[main] [16:2:27:829] [GSDCTLDriver.execute:77] PRKC-1021 : Problem in the

clusterware

PRKC-1021 : Problem in the clusterware

Failed to get list of active nodes from clusterware

[main] [16:2:27:832] [line# N/A] PRKC-1021 : Problem in the clusterware

[main] [16:2:27:832] [line# N/A] exiting abnormally due to FrameworkException

17-OCT-05 09:27:04 GMT

New info : AHDXJF : I will update to 9206 .

But when i install patch 9206, This patch could not find two nodes to update.

Does not appear two nodes install step.

So,What can i do

ORACLE到现在还没有答复，大家看看到底是OS的问题，还是HA的问题还是ORACLE的问题？要打什么PATCH？