[ovs-dev] [PATCH v2] IntegrationGuide: A guide to help integrators.

[Ben Pfaff]

On Wed, Jun 20, 2012 at 04:53:35PM -0700, Justin Pettit wrote:
> 
> On May 17, 2012, at 12:37 PM, Ben Pfaff wrote:
> 
> > On Thu, May 17, 2012 at 12:01:45PM -0700, Justin Pettit wrote:
> >> Thanks for writing this up, Guru!  Since we do have integrators that
> >> are not porting this to hypervisors, I took a crack at making it
> >> less hypervisor-specific.  I've added my version below.  What do you
> >> think?
> >> 
> >> By the way, we're accumulating a lot of docs in the root directory.
> >> What do people think about moving them to a "docs" directory or
> >> something?  Also, we use a number of number of different
> >> capitalization styles.  If this is the preferred style, should we
> >> think about renaming ones like "PORTING" to "Porting"?
> > 
> > This is a bad example:
> >>  ovs-vsctl set Bridge br0 other_config:hw_addr="01:23:45:67:89:0a"
> > First, because it's "hwaddr" not "hw_addr".  Second, because that is
> > an invalid MAC address (the multicast bit is set).
> 
> Fixed.
> 
> > The description of "attached-mac" has some general information about
> > how hypervisors typically work that might be promoted up a level.  It
> > also doesn't mention that the host-side MAC is often not unique
> > (XenServer uses a fixed value of FE:FF:FF:FF:FF:FF, I think).
> 
> Okay, it should be cleaned up a bit.
> 
> > I recently added a carefully worded description of iface-status to
> > ovs-vswitchd.conf.db(5).  If you didn't already read it, it might be
> > worthwhile, to see if one description or the other might be improved.
> 
> I just opted to provide a pointer to the ovs-vswitchd.conf.db man page.
> 
> > On an editorial, there seems to be a lot more passive voice than
> > necessary here.
> 
> Okay, should be better now.
> 
> I've updated a revised version below.

I'm happy with this based on a quick skim.  Thank you for the updates!

Ben