I want to know what would be the best way to uncover a library through jerom. Say, I install a machine learning library (MLL) on a machine, and I have a zoromic broker running on the other. Now, if I have a Zorom client who needs to call the function within ML, then how can I do this through a broker?
I want to know this work that what I have to do to do this work. In general, you should be a "listener" that raises the data from ZMQ.
and feeds it on your machine-learning backend code, then transmits the results back to the requester is.
A lot of designs are preferred, such as serialing of data between client and data server (JSON? YAM? Real? Saving? ...), and how to specify requests and request options But all things are considered, this is a very straightforward ZMQ use.
The problem occurs when you have a more feature-rich, full, robust, etc. Design - things like multi-threaded or multi-process server, multi-machine scalability, secure user / request authentication and authorization, job reporting And dashboard, or job checkpointing. These are all "extra" normal "network scheduler" or "enterprise message brokers" functions that come with or along the package-like-out-the-box.
If you do not want to go through the full "message broker middleware" route, you can start by checking for the design of others for lightweight ZMQ-based job brokers, like
No comments:
Post a Comment