Hi,

Ignite distributed messaging use the same of mechanism that Ignite uses for
all internal messages. Ignite sends user messages to nodes in cluster by
TcpCommunicationSpi. On remote nodes messages will be processed in public
pool. Sender and Listener fully decoupled and processing speed in listener
doesn't affect the speed of send method.

On Wed, Apr 26, 2017 at 6:18 PM, ght230 <[email protected]> wrote:

> I have one application using Ignite distributed messaging.In this
> application, Ignite receive messages from outside and sendordered them by
> Ignite distributed messaging. On the other side, after Ignite listener
> receive these messages, it will process them.I have recorded the time it
> took from the message received from outside to the message sendordered.
>
> I have run this application in 2 different enviroment.
> In the first enviroment, this will take a few milliseconds.
> In the second envirement, normally this will take a few milliseconds.But
> sometimes it will take hundreds of milliseconds.
> The second enviroment's hardware configuration is not as good as the first
> one.
>
> I want to know what is mechanism of Ignite distributed messaging?
> Are sender and the listener fully decoupled?
> Is there backpressure mechanism in Ignite distributed messaging?
> Will the processing speed in the listener affect the speed of sendordered?
>
>
>
> --
> View this message in context: http://apache-ignite-users.
> 70518.x6.nabble.com/What-is-the-mechanism-of-Ignite-
> distributed-messaging-tp12272.html
> Sent from the Apache Ignite Users mailing list archive at Nabble.com.
>

Reply via email to