[rabbitmq-discuss] Beginner Question.

[ratheesh kannoth]

I am going thru some documentation on rabbitmq. I have some questions


Client A ( Running on Machine A )   would like to send a message to
Client B ( running on machin B ). Broker  ( Rabbit Mq server ) is
running on Machine C.

1. So client A using rabbitmq library  makes a connection to Broker C
over tcp/ip and request for xchange and queue. Is my understanding
correct ?

2. If client A crashes, messages reliablility   is guranteed by Broker
C.   Durability will ensure messages are kept during broker restart.
How can we guarantee messages are retained upon a crash ( not a
graceful shutdown ) of a broker.   my use case here :  Suppose I  have
 10 nodes connected on a lan. These nodes need to send messages to
each other. So i need to run broker ( rabbitmq server ) on one node.
I don't have budget to run one node dedicated to broker ; so i will
run some of my applications also on that node. These applications may
crash the broker node ( can make  the node reboot  ).  what is the
solution here ?

3. How is rabbitmq C is  well tested ? or it is in development phase.
I could see a lot of examples on java etc. But in C , it is limited.?


Thanks,
Ratheesh