openstack team mailing list archive

[韦远科 <weiyuanke123@xxxxxxxxx>]

thanks gong and sandy for your brief and clear explanation.  really helps

-----------------------------------------------------------------
韦远科
中国科学院 计算机网络信息中心
北京市海淀区中关村南四街四号 349信箱(100190)

Yuanke Wei (Wei)
Computer Network Information Center,Chinese Academy of Sciences
4 Zhongguancun Nansijie, Haidian District, Beijing 100190, China
P.O. Box 349
gtalk: weiyuanke123@xxxxxxxxx
msn: weiyuanke123@xxxxxxxxxxx



On Tue, Oct 30, 2012 at 7:43 PM, gong yong sheng <gongysh@xxxxxxxxxxxxxxxxxx
> wrote:

>  Following is my understanding:
> usage:
> scheduler queue is used for compute api server. For example of nova boot,
> API server will send a msg to scheduler queue for a booting
> scheduler_*fanout_xxxx is used for services to report service capability.
> For example, nova compute *will use scheduler_fanout to publish its
> capabilities.
>
> scheduler.{hostname} seemingly is not used now.
> we can deploy many schedulers into a Openstack cloud. All the schedulers
> will listen on scheduler msg queue. AMQP server (rabbitmq) will adopt round
> robin way to
> dispatch the msg to these schedulers. But since the capabilities are fan
> outed, all the schedulers can receive these capabilities.
>
>
> On 10/30/2012 06:55 PM, 韦远科 wrote:
>
> hi all,
>
>  I read into the source code for nova-scheduler and found there actually
> exists three msg queues, like
> "scheduler"
> "scheduler.node70"
> "scheduler_*fanout_bd738fedcdf344d9bb3cb580657f54e0*".
>
>  what's the functions for each queue and are there connections??
>
>
>  thanks,
>
>
>
>
>
>
> _______________________________________________
> Mailing list: https://launchpad.net/~openstack
> Post to     : openstack@xxxxxxxxxxxxxxxxxxx
> Unsubscribe : https://launchpad.net/~openstack
> More help   : https://help.launchpad.net/ListHelp
>
>
>