[rabbitmq-discuss] question : how to speed up the worker ?
bino oetomo
bino at indoakses-online.com
Thu May 17 03:54:23 BST 2012
Dear All.
On 05/16/2012 05:15 PM, Tim Watson wrote:
>> Is there any potential conflict/problem when I use multi-worker that
>> can do 'reject' ?
>>
>
>
> Not really. Once a message has been delivered to a worker, it won't be
> seen by any other worker unless it is rejected and re-queued or the
> worker dies without sending an ACK. HTH.
>
> Tim
>
>
Since my concern about worker speed is caused by the fact that the fleet
will only have a short time to transfer it's local data to the 'app server',
i try to find another work around , and I think i will deploy Sholve plugin.
That is :
1. Fleet Side
a. Sender will send to local broker
b. Sholve will handle the transfer to remote-broker (located in a
controll room)
2. Controll room side :
a. Broker will 'receive' msg from Fleet-Sholve
b. Worker will read and injecting the msg to Application-server.
Hopefully those scheme will :
A. Speed up the data unloading from fleet to server, and keep safe the
process.
B. When I need to speed up the worker, I can do it on a single Engine.
No need to re-config the fleet-side.
so All .....Kindly please give me your opinion about those scheme.
C/q : Tim and Alvaro, I really appreciate your enlightment.
Sincerely
-bino-
More information about the rabbitmq-discuss
mailing list