Cluster Single Node Failure (no replies)

I deploy a mysql cluster7.3.5 on windows(1 management node, 2 data nodes, 2 sql nodes) on 2 machines:
Management Nde: Machine A
Data node A: Machine A
Sql Node A: Machine A

Data node B: Machine B
Sql Node B: Machine B

I do following actions:
1. Create a Database named ClusterSample with Sql Node A.
2. Create a tabale named Person with NDB engine for the database ClusterSample with Sql Node A
3. Insert into a record into Table Person.
4. query the record with Sql Node B, pass.
5. Disable the network of the Machine A.
6. query the record with sql Node B again. We get an error on the sql node: "Incorrect information in file:'.\ClusterSample\Person.frm'" and failed to get query result. the data node process is terminated unexpectly.
get following log on data node:
Status: Temporary error, restart node
Message: Node lost connection to other nodes and can not form a unpartitioned cluster, please investigate if there are error(s) on other node(s) (Arbitration error)
Error: 2305
Error data: Arbitrator decided to shutdown this node
Error object: QMGR (Line: 6043) 0x00000002

my question is: Why after the machine A is disconnected, the sql node on machine B also can't work correctly. in my understanding, the sql node on machine B can work correctly since the single node failure won't affect the whole cluster.