Open Stack

On 5/25/2016 13:24, Tim Rozet wrote:
> In my opinion, it is a better approach to break this down into plugin vs driver support.  There should be no problem adding support into networking-sfc plugin for NSH today.  The OVS driver however, depends on OVS as the dataplane - which I can see a solid argument for only supporting an official version with a non-NSH solution.  The plugin side should have no dependency on OVS.  Therefore if we add NSH SFC support to an ODL driver in networking-odl, and use that as our networking-sfc driver, the argument about OVS goes away (since neutron/networking-sfc is totally unaware of the dataplane at this point).  We would just need to ensure that API calls to networking-sfc specifying NSH port pairs returned error if the enabled driver was OVS (until official OVS with NSH support is released).
>

Does ODL have a dataplane? I thought it used OvS. Is the ODL project 
supporting its own fork of OvS that has NSH support or is ODL expecting 
that the user will patch OvS themself?

I don't know the details of why OvS hasn't added NSH support so I can't 
judge the validity of the concerns, but one way or another there has to 
be a production-quality dataplane for networking-sfc to front-end.

If ODL has forked OvS or in some other manner is supporting its own NSH 
capable dataplane then it's reasonable to consider that the ODL driver 
could be the first networking-sfc driver to support NSH. However, we 
still need to make sure that the API is an abstraction, not 
implementation specific.

But if ODL is not supporting its own NSH capable dataplane, instead 
expecting the user to run a patched OvS that doesn't have upstream 
acceptance then I think we would be building a rickety tower by piling 
networking-sfc on top of that unstable base.