I have a brand new setup with zabbix 1.8.4 with Postgresql.
I also using CentOS 5.5 Fresh Clean Install as the host OS.
Both machines are exactly the same setup.
After some hours the setup works pretty well as standalone config, so both Nodes are NodeID=0.

Then I setup a distributed monitoring like the following:

Node 10=172.22.0.10, Master
Node 20=172.23.0.10, Slave

I setup the nodes in both Web-Interfaces, see screenshot (masternode.png and slavenode.png).
I also attached zabbix_server.log of the master node which is crashing and a objdump of the zabbix_server file.