Re: [Xen-users] remus network buffering configuraion

[张伟 <zhangwqh@xxxxxxx>]

Thank you for your advice! 

I have contacted with Shriram and he said that my following understanding is right. 

The implementation of sch_plug is based on fifo. It does not support class, so we can not use it to do some filter settings. If we only want to buffer some specific packets(ip packets, or ip dst, ip src, mac dst, mac src and so on), we have to rely on a middle device. ingress qdisc can do some filter settings, then redirect the packets to this middle device -ifb.

在 2014-01-06 18:41:42，"Ian Campbell" <Ian.Campbell@xxxxxxxxxx> 写道：
>On Mon, 2013-12-23 at 05:32 +0800, 张伟 wrote:
>> Hi all,
>> 
>> 
>> Does anybody know that the function of ifb in xen remus project? I can
>> not understand why remus network buffering needs ifb device. In new
>> xen and linux version, the kernel qdisc module sch_plug provides the
>> network buffering. If a virtual machine - eg domain id is 1. In
>> domain0, vif is vif1.0(tc qdisc add dev vif1.0 root plug). Why not
>> directly use the plug qdisc on this vif1.0. Remus needs to  use ifb
>> device. The packets of vif1.0 needs to redirect to ifb0. 
>
>I think you should ask (by CCing) the Remus maintainer, as listed in the
>MAINTAINERS file. I've added Shriram for you this time.
>
>> 
>> 
>> ip link set ifb0 up
>> 
>> 
>> tc qdisc add dev ifb0 root plug
>> 
>> 
>> tc qdisc add dev vif1.0 ingress
>> 
>> 
>> tc filter add dev  vif1.0  parent ffff: proto ip pref 10 u32 match u32
>> 0 0 action mirred egress redirect dev ifb0
>> 
>> 
>> 
>> 
>> _______________________________________________
>> Xen-users mailing list
>> Xen-users@xxxxxxxxxxxxx
>> http://lists.xen.org/xen-users
>
>
>
>_______________________________________________
>Xen-users mailing list
>Xen-users@xxxxxxxxxxxxx
>http://lists.xen.org/xen-users

_______________________________________________
Xen-users mailing list
Xen-users@xxxxxxxxxxxxx
http://lists.xen.org/xen-users