Programs like people need to communicate, and for them we have the UDP, TCP, HTTP, IPX, WebSocket protocol to connect and other related applications. But the underlying protocol is difficult to achieve.
We need a high level of abstraction, scalable, and easy to use things, it is ZeroMQ(ØMQ). ØMQ gives us advanced levels of availability and speed.
ØMQ is a neat messaging library that allows us to build our own messaging infrastructure. It can help build framework that scales where services could be handled by different applications.
I am inclined to python programming and luckily pyzmq provides python bindings for ØMQ and allows us to leverage ØMQ in python applications.
This tutorial is my initial journey into ØMQ programming with python.
Lets start with ØMQ Messaging Patterns:-
In distributed architecture, different parts of system interconnect and communicate with each other. These interconnecting systems viewed graphically represents a network topology. Messaging patterns are network oriented architectural pattern that describes the flow of communication between interconnecting systems. ØMQ provides pre-optimized sockets which enables you to take advantage of these patterns.
Each pattern in ØMQ defines the constraints on the network topology. What systems can connect to each other and flow of communication between them. These patterns are designed to scale.
Here show you each of the pattern and PART 2(releasing soon) with example code for each pattern.
It provides sockets that are close in behavior to conventional sockets.
Conventional sockets allow:
• only strict one-to-one (two peers)
• many-to-one (many clients, one server)
• one-to-many (multicast) relationships
Client / Server:-
Most basic pattern is client/server model, where client sends a request and server replies to the request.
There is one difference from zmq.PAIR and other type of ZMQ sockets.
• ZMQ REQ sockets can connect to many servers.
• The requests will be interleaved or distributed to both the servers.
With socket zmq.PAIR, you could send any number of messages among connected peers or client/server.
• socket zmq.REQ will block on send unless it has successfully received a reply back.
• socket zmq.REP will block on recv unless it has received a request.
Each Request/Reply is paired and has to be successful.
Publish/Subscribe is another classic pattern where senders of messages, called publishers, do not program the
messages to be sent directly to specific receivers, called subscribers. Messages are published without the knowledge
of what or if any subscriber of that knowledge exists.
Push and Pull sockets let you distribute messages to multiple workers, arranged in a pipeline. A Push socket will
distribute sent messages to its Pull clients evenly. This is equivalent to producer/consumer model but the results
computed by consumer are not sent upstream but downstream to another pull/consumer socket.
Data always flows down the pipeline, and each stage of the pipeline is connected to at least one node. When a
pipeline stage is connected to multiple nodes data is load-balanced among all connected nodes.
Continue on Part2 (ØMQ Devices, Polling and Sockets)…
download link: http://download.zeromq.org/