On Mon, Apr 20, 2015 at 03:45:09PM -0400, Russell Bryant wrote:
> On 04/18/2015 01:04 PM, Ben Pfaff wrote:
> > It's possible that we could implement all of the OVN concepts as a single
> > collection of flow tables, but it makes more sense to me if we instead
> > represent each logical switch and logical router separately. This also
> > fits the model described in T. Koponen et al., "Network Virtualization in
> > Multi-tenant Data Centers," NSDI '14, the model that OVN aspires to
> > imitate.
> >
> > Signed-off-by: Ben Pfaff <b...@nicira.com>
> > ---
> > ovn/ovn-sb.ovsschema | 2 ++
> > ovn/ovn-sb.xml | 16 +++++++++++++++-
> > 2 files changed, 17 insertions(+), 1 deletion(-)
> >
> > diff --git a/ovn/ovn-sb.ovsschema b/ovn/ovn-sb.ovsschema
> > index 98662b8..5f2d1a4 100644
> > --- a/ovn/ovn-sb.ovsschema
> > +++ b/ovn/ovn-sb.ovsschema
> > @@ -33,6 +33,7 @@
> > "max": "unlimited"}}}},
> > "Pipeline": {
> > "columns": {
> > + "logical_datapath": {"type": "uuid"},
> > "table_id": {"type": {"key": {"type": "integer",
> > "minInteger": 0,
> > "maxInteger": 127}}},
> > @@ -44,6 +45,7 @@
> > "isRoot": true},
> > "Bindings": {
> > "columns": {
> > + "logical_datapath": {"type": "uuid"},
> > "logical_port": {"type": "string"},
> > "parent_port": {"type": {"key": "string", "min": 0, "max":
> > 1}},
> > "tag": {
> > diff --git a/ovn/ovn-sb.xml b/ovn/ovn-sb.xml
> > index e4159db..45b8622 100644
> > --- a/ovn/ovn-sb.xml
> > +++ b/ovn/ovn-sb.xml
> > @@ -227,6 +227,13 @@
> > The default action when no flow matches is to drop packets.
> > </p>
> >
> > + <column name="logical_datapath">
> > + The logical datapath to which the logical port belongs. A logical
> > + datapath implements a logical pipeline among the ports in the <ref
> > + table="Bindings"/> table associated with it. (No table represents a
> > + logical datapath.)
> > + </column>
> > +
> > <column name="table_id">
> > The stage in the logical pipeline, analogous to an OpenFlow table
> > number.
> > </column>
> > @@ -446,7 +453,8 @@
> > they are Unicode strings). String constants are used for naming
> > logical ports. Thus, the useful values are <ref
> > column="logical_port"/> names from the <ref column="Bindings"/> and
> > - <ref column="Gateway"/> table.
> > + <ref column="Gateway"/> table in a logical flow's <ref
> > + column="logical_datapath"/>.
> > </p>
> >
> > <p>
> > @@ -653,6 +661,12 @@
> > <code>chassis</code> column with new information.
> > </p>
> >
> > + <column name="logical_datapath">
> > + The logical datapath to which the logical port belongs. A logical
> > + datapath implements a logical pipeline via logical flows in the <ref
> > + table="Pipeline"/> table. (No table represents a logical datapath.)
> > + </column>
> > +
> > <column name="logical_port">
> > A logical port, taken from <ref table="Logical_Port" column="name"
> > db="OVN_Northbound"/> in the OVN_Northbound database's
> >
>
> This all looks good to me, so:
>
> Acked-by: Russell Bryant <rbry...@redhat.com>
>
> I wonder if it's worth specifically drawing the connection between the
> logical datapath in ovn-sb and the logical switch in ovn-nb. I see in
> the code later that it's set as the logical switch UUID from ovn-nb.
> From a schema perspective, I get that it just needs to be a UUID. Maybe
> an update to ovn-architecture could mention that ovn-northd creates
> these entries using the logical switch's UUID to tie them together.
Thanks. I tried to do this, although I'm not sure I did a great job.
See my reply to Justin for the diff.
_______________________________________________
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/listinfo/dev
