Hi,
I've read chapter 1 & 2 of the Guide so far, and some posts in the mailing
list.
I'm wondering about the following case:
A client C asks a server S about the status of the workers W(1..n).
(the number of workers can vary)
Here is the implementation I've come up with in terms of communication
patterns:
C:REQ -> REP:S
S:PUB -> SUB:W1..Wn
W1..Wn:PUSH -> PULL:S
S:REP -> REQ:C
- C sends a REQ to S
- S sends the status request through PUB to the workers, who have
subscribed to a 'broadcast' channel.
- the workers send their answers through PUSH-PULL to the server
- the server replies with REP to the client.
My main problem is: how do I figure out whether I have every answer back?
Or asked differently, is this the right architecture for the problem I
want to solve?
For now, I count the answers coming in through the PULL socket, and when I
reach the number of known workers, I wrap up my answer to the client.
It seems to work at first. But later on, I get answers that look like
replies to earlier requests, so something got out of step.
Any suggestion?
--
Empfehlen Sie GMX DSL Ihren Freunden und Bekannten und wir
belohnen Sie mit bis zu 50,- Euro! https://freundschaftswerbung.gmx.de
_______________________________________________
zeromq-dev mailing list
[email protected]
http://lists.zeromq.org/mailman/listinfo/zeromq-dev
