ndbd --initial

I run MySQL Cluster with

1 MGM node 2 DB nodes and 1 API node

When MySQL Cluster is running 2 DB nodes crashed!

with config.ini file:

[COMPUTER DEFAULT] [DB DEFAULT] NoOfReplicas: 2

[MGM DEFAULT] ArbitrationRank: 2

[TCP DEFAULT] PortNumber: 28002

[COMPUTER] Id: 1 HostName: 10.1.29.101

[COMPUTER] Id: 2 HostName: 10.1.29.102

[COMPUTER] Id: 3 HostName: 10.1.29.103

[COMPUTER] Id: 5 HostName: 10.1.29.105

[MGM] Id: 1 ExecuteOnComputer: 1 PortNumber: 28000

[DB] Id: 2 ExecuteOnComputer: 2 FileSystemPath: /usr/local/mysql_cluster/db-2 IndexMemory: 1M DataMemory: 1M

[DB] Id: 5 ExecuteOnComputer: 5 FileSystemPath: /usr/local/mysql_cluster/db-5 IndexMemory: 1M DataMemory: 1M

[API] Id: 3 ExecuteOnComputer: 3

1st DB node crashed with error: **************************************** Date/Time: Tuesday 21 September 2004 - 16:26:09 Type of error: error Message: Invalid Configuration fetched from Management Server Fault ID: 2350 Problem data: Illegal configuration change. Initial start needs to be performed when changing no of replicas (1 != 2) Object of reference: DBDIH (Line: 3508) 0x0000000a ProgramName: NDB Kernel ProcessID: 386 TraceFile: ./ndb_2_trace.log.8

2nd DB node crashed with error: (when NoOfReplicas=2) ***************************************** Date/Time: Tuesday 21 September 2004 - 16:26:16 Type of error: error Message: Internal program error (failed ndbrequire) Fault ID: 2341 Problem data: QmgrMain.cpp Object of reference: QMGR (Line: 1833) 0x0000000a ProgramName: NDB Kernel ProcessID: 390 TraceFile: ./ndb_5_trace.log.6 ***EOM***

or 2nd DB node crashed with error: (when NoOfReplicas=1) ****************************************** Date/Time: Tuesday 21 September 2004 - 16:41:30 Type of error: error Message: Arbitrator shutdown Fault ID: 2305 Problem data: Arbitrator decided to shutdown this node Object of reference: QMGR (Line: 3774) 0x0000000a ProgramName: NDB Kernel ProcessID: 489 TraceFile: ./ndb_5_trace.log.7

please help!