one ndb note cant start and it said "Internal program error (failed ndbrequire)"

[Li Ma]

two machines, both run as sql node + ndb node.
doing "alter table engine=ndbcluster": in the beginning, all seem ok, no error report, until it reports there has node failure.

mysql> alter table members engine=ndbcluster;
Query OK, 223682 rows affected (8 min 16.52 sec)
Records: 223682 Duplicates: 0 Warnings: 0

mysql> alter table posts engine=ndbcluster;
ERROR 1297 (HY000): Got temporary error 4031 'Node failure caused abort of transaction' from ndbcluster

tried to restart ndb manager and ndb notes and sql notes, but one note still can't work.

############################# the error log is : ##############################
Time: Monday 28 August 2006 - 20:52:41
Status: Temporary error, restart node
Message: Internal program error (failed ndbrequire) (Internal error, programming error or missing error message, please report a bug)
Error: 2341
Error data: Dbdict.cpp
Error object: DBDICT (Line: 1080) 0x0000000a
Program: ndbd
Pid: 23896
Trace: /usr/local/mysql/data/ndb_2_trace.log.19
Version: Version 5.0.24
***EOM***

this error ever appeared in early time, but when I restarted ndbd and ndb_mgmd, it can work. this time, I tried and tried to restart the service, but it still can't work.
########################### the cluster config.ini is: ##########################
# Options affecting ndbd processes on all data nodes:
[NDBD DEFAULT]
NoOfReplicas=2
DataMemory=2000M
IndexMemory=200M

MaxNoOfAttributes=10000
MaxNoOfTables=2000
MaxNoOfUniqueHashIndexes=5000
MaxNoOfOrderedIndexes=5000
MaxNoOfTriggers=4000
TimeBetweenLocalCheckpoints=2
NoOfFragmentLogFiles=100

# TCP/IP options:
[TCP DEFAULT]
portnumber=2202

# Management process options:
[NDB_MGMD]
id = 1
hostname=192.168.100.225
datadir=/usr/local/mysql-cluster

# Options for data node "A":
[NDBD]
id = 2
hostname=192.168.100.206
datadir=/usr/local/mysql/data

# Options for data node "B":
[NDBD]
id = 3
hostname=192.168.100.225
datadir=/usr/local/mysql/data

# SQL node options:
[MYSQLD]
id = 4
hostname=192.168.100.206

# SQL node options:
[MYSQLD]
id = 5
hostname=192.168.100.225

in the failed ndb note:
free
total used free shared buffers cached
Mem: 1533604 159568 1374036 0 2968 125232
-/+ buffers/cache: 31368 1502236
Swap: 2096472 15112 2081360