***SPAM*** Re: [ofa-general] RE: QoS setting and propagation

arkady kanevsky arkady.kanevsky at gmail.com
Fri Apr 3 09:30:16 PDT 2009


Caitlin,
can you clarify what you are suggesting?
Are you proposing that RDMA APIs for QoS setting is the not the way to go?
Are you proposing to use netdev mechanisms for setting RFC2474 DS fields to
be used for RDMA QoS setting?
Something else?
Thanks,
Arkady

On Fri, Apr 3, 2009 at 12:13 PM, Caitlin Bestler
<caitlin.bestler at gmail.com>wrote:

>  On Thu, Apr 2, 2009 at 5:28 PM, Sean Hefty <sean.hefty at intel.com> wrote:
> >>So the plan is to hook up service_type/tos to VLAN and skb->priority
> >>for iWARP? But since we do not setup socket connection explicitly
> >>we can not use SO_PRIORITY field to do it. Is this correct?
> >>Do we have a plan on how to hook up tos without socket?
> >
> > I really don't have any plans at this time to link QoS and iWarp.
>  Someone
> > working closer with iWarp would need to provide any sort of
> implementation.
> >
> Implementing a parallel solution for VLAN/priority and other policy
> controls
> for iWARP connections would be terribly wasteful, both in development time
> and in the need for network administrators to use two different sets of
> tools
> to implement what should be an integrated set of policies.
>
> But that would require the netdev stack to acknowledge the existence of
> offloaded connections, and instead of just waiting for them to go away
> insist
> that they follow a minimal set of rules to provide co-ordinated management.
>
> In my opinion this won't happen because vendors push for it. Fabric users
> are going to have to make their need for uniform control of
> VLAN/priorities/etc.
> understood.
>



-- 
Cheers,
Arkady Kanevsky
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openfabrics.org/pipermail/general/attachments/20090403/15d56d85/attachment.html>


More information about the general mailing list