[ovs-dev] [ovn-controller-vtep V4 2/6] ovn-sbctl: Add ovn-sbctl.
Alex Wang
alexw at nicira.com
Thu Jul 16 16:06:24 UTC 2015
On Thu, Jul 16, 2015 at 6:38 AM, Russell Bryant <rbryant at redhat.com> wrote:
> On 07/16/2015 03:56 AM, Alex Wang wrote:
> > +static void
> > +usage(void)
> > +{
> > + printf("\
> > +%s: ovs-vswitchd management utility\n\
> > +\n\
> > +for debugging and testing only, never use it in production\n\
> > +\n\
> > +usage: %s [OPTIONS] COMMAND [ARG...]\n\
> > +\n\
> > +SouthBound DB commands:\n\
> > + show print overview of database contents\n\
> > +\n\
> > +Chassis commands:\n\
> > + add-chassis CHASSIS create a new chassis named CHASSIS\n\
> > + del-chassis CHASSIS delete CHASSIS and all of its encaps,\n\
> > + and gateway_ports\n\
> > +\n\
> > +Binding commands:\n\
> > + bind-lport LPORT CHASSIS bind logical port LPORT to CHASSIS\n\
> > + unbind-lport LPORT delete the binding of logical port
> LPORT\n\
>
> In ovn-nbctl, we have a convention of starting every command name with
> the resource you're operating on, followed by the action you're taking.
> For example:
>
>
> Logical switch commands:
> lswitch-add [LSWITCH] create a logical switch named LSWITCH
> lswitch-del LSWITCH delete LSWITCH and all its ports
> lswitch-list print the names of all logical switches
> lswitch-set-external-id LSWITCH KEY [VALUE]
> set or delete an external-id on LSWITCH
> lswitch-get-external-id LSWITCH [KEY]
> list one or all external-ids on LSWITCH
>
>
> In this patch, you've taken a different approach. I remember debating
> whether we should change ovn-nbctl (with Justin I believe), with the
> primary argument for changing it being to more closely match the style
> of ovs-vsctl. However, I still like the convention in ovn-nbctl better.
> :-)
>
>
Yeah, I followed the ovs-vsctl convention, of course~ I'll change to
follow the
ovn convention~ ;D
> So, I'd propose we adopt the same convention in ovn-sbctl that ovn-nbctl
> uses. The specific changes would be:
>
> add-chassis -> chassis-add
> del-chassis -> chassis-del
> bind-lport -> lport-bind
> unbind-lport -> lport-unbind
>
> Otherwise, the patch looks good to me.
>
> --
> Russell Bryant
>
More information about the dev
mailing list