[ovs-dev] [PATCH] ovn-northd: create patch ports when necessary.

[nickcooper-zhangtonghao]

I only consider the OVN without CMS. Your tips may sound great. 

> On Jul 29, 2016, at 10:48 AM, Ryan Moats <rmoats at us.ibm.com> wrote:
> 
> "dev" <dev-bounces at openvswitch.org> wrote on 07/27/2016 09:47:47 PM:
> 
> > From: nickcooper-zhangtonghao <nickcooper-zhangtonghao at opencloud.tech>
> > To: Ben Pfaff <blp at ovn.org>
> > Cc: dev at openvswitch.org
> > Date: 07/27/2016 09:48 PM
> > Subject: Re: [ovs-dev] [PATCH] ovn-northd: create patch ports when necessary.
> > Sent by: "dev" <dev-bounces at openvswitch.org>
> > 
> > 
> > > On Jul 28, 2016, at 5:03 AM, Ben Pfaff <blp at ovn.org> wrote:
> > > 
> > > On Mon, Jul 25, 2016 at 09:23:19AM -0700, nickcooper-zhangtonghao wrote:
> > >> If the logical router ports without 'peer' or the port named peer
> > >> is not created, it is unnecessary to insert the ports into the
> > >> southbound Port_Binding table.
> > >> 
> > >> Similarly, if logical switch ports of type 'router' don't contain
> > 'router-port'
> > >> value, these ports will not be inserted.
> > >> 
> > >> This patch may optimize the process of 'build_ports'.
> > >> 
> > >> Signed-off-by: nickcooper-zhangtonghao <nickcooper-
> > zhangtonghao at opencloud.tech>
> > > 
> > > This doesn't seem like a common case.  Is it worth optimizing?
> > 
> > It seems to me that it’s unnecessary to create the ports unused currently.
> 
> If I think about it, the router port and its peer will be created by the
> CMS.  That means this would optimize for the case where the CMS creates
> these two ports at substantially different times.  I can't think of any
> CMS that works that way today, so I'm with Ben, this doesn't sound like a
> common enough case.
>