Hey Marak,

Thanks for the valuable info.
The use case is: I have a few hooks that receive messages each one from a different source (rabbitMQ, http) and there is a central hook that receives all those messages, does some calculations and stores into a database.

Truth is, our message rate is quite lower than 600 msg/s but i need to pass in some initial data which are ~30mil, so that means i will have to write separate code to do just this, without hook.io.

I could also pack all that stuff into one process and get over it, but then i wouldn't make use of multicore etc. Do you know other communication/IPC libraries i could look into, or got any other suggestion?

Thanks,
danmilon.

On 07/02/2012 01:35 AM, Marak Squires wrote:
As of today, I've done literally no performance optimizations for hook.io <http://hook.io>.

hook.io <http://hook.io>'s power is in API and discovery.

If you require more then 666.66 messages per second out of the box, you should not be using dnode as your message transport. hook.io <http://hook.io> currently uses dnode as it's primary message transport.

hook.io <http://hook.io> is intended to be multi-transport, so for high performance situations you'd want to use hook.io <http://hook.io> as a wrapper around a more robust solution like 0mq.

If do you don't mind, I'd be interested in hearing what your intended use case for hook.io <http://hook.io> was which would require that throughput. Always good to get more input to shape the API.

On Sun, Jul 1, 2012 at 3:18 PM, Dan Milon <danmi...@gmail.com <mailto:danmi...@gmail.com>> wrote:

    Hello,

    I am using hook.io <http://hook.io> for some project.
    Today i thought of measuring how fast hook.io <http://hook.io> can
    emit and receive messages, so i built a dead simple benchmark [1],
    and numbers are not that interesting.
    It took 15 secs to deliver 10k messages from one hook to another.

    I do not know the underlaying mechanisms hook.io <http://hook.io>
    uses for IPC (please enlighten me), but isn't this performance
    quite low? Or is there something wrong with the test?

    Thanks,
    Dan Milon.

    [1] https://gist.github.com/3029842

-- Job Board: http://jobs.nodejs.org/
    Posting guidelines:
    https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines
    You received this message because you are subscribed to the Google
    Groups "nodejs" group.
    To post to this group, send email to nodejs@googlegroups.com
    <mailto:nodejs@googlegroups.com>
    To unsubscribe from this group, send email to
    nodejs+unsubscr...@googlegroups.com
    <mailto:nodejs%2bunsubscr...@googlegroups.com>
    For more options, visit this group at
    http://groups.google.com/group/nodejs?hl=en?hl=en


--
Job Board: http://jobs.nodejs.org/
Posting guidelines: https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines
You received this message because you are subscribed to the Google
Groups "nodejs" group.
To post to this group, send email to nodejs@googlegroups.com
To unsubscribe from this group, send email to
nodejs+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/nodejs?hl=en?hl=en


--
Job Board: http://jobs.nodejs.org/
Posting guidelines: 
https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines
You received this message because you are subscribed to the Google
Groups "nodejs" group.
To post to this group, send email to nodejs@googlegroups.com
To unsubscribe from this group, send email to
nodejs+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/nodejs?hl=en?hl=en

Reply via email to