Hi All,
In packet_out message when we specify the OFPP_TABLE reserved port as an
output action, then actions are perform in flow tables on packet.
There are 254 tables, my question is that who's actions will be perform on
it out of 254.
Regards,
Manpreet Singh
__
On Wed, Sep 23, 2015 at 09:11:26PM +0200, Daniel Herrmann wrote:
> I am trying to load-balance traffic in a routed scenario and want to use
> OF group tables for this.
>
> However, it seems that only one bucket is selected. I found some thread
> from mid 2014, which states that only eth_dst is use
On Wed, Sep 23, 2015 at 11:31:31AM +, Daniel Swahn wrote:
> (gdb) frame 0
> #0 0x004291ca in rstp_run (ofproto=,
> ofproto=) at ofproto/ofproto-dpif.c:2278
> 2278bundle_move(((struct ofport_dpif
> *)rstp_get_old_root_aux(ofproto->rstp))->bundle,
Thanks for the report
I see. thanks.
+
Best,
Hao
On Tue, Sep 29, 2015 at 11:34 AM, Ben Pfaff wrote:
> On Tue, Sep 29, 2015 at 11:33:15AM +0800, Hao Wu wrote:
> > I think each megaflow table is built according to a certain tuple
> > consisting of fields. If we don't care some bits in the packet, we
On Tue, Sep 29, 2015 at 11:33:15AM +0800, Hao Wu wrote:
> I think each megaflow table is built according to a certain tuple
> consisting of fields. If we don't care some bits in the packet, we won't
> select them in the matching tuple, so why we still need the wildcard bits?
Which bits a flow ma
Hi Ben,
I think each megaflow table is built according to a certain tuple
consisting of fields. If we don't care some bits in the packet, we won't
select them in the matching tuple, so why we still need the wildcard bits?
+
Best,
Hao
On Tue, Sep 29, 2015 at 11:16 AM, Ben Pfaff
On Mon, Sep 28, 2015 at 04:52:01AM +, Xiao Liang (xiaolia) wrote:
> Is it possible to add negative match support in OpenFlow or as an
> extension?
The main problem is to design an efficient implementation of negative
match in the classifier. I don't know a way to do it in a general way.
_
On Tue, Sep 29, 2015 at 11:12:17AM +0800, Hao Wu wrote:
>In OVS NSDI'15 paper, it mentions there is a new megaflow cache in the
> kernel space. I think it is an active subset of the flow table in user
> space. Does a megaflow entry contain any wildcard bits?
Yes.
__
Hi all,
In OVS NSDI'15 paper, it mentions there is a new megaflow cache in the
kernel space. I think it is an active subset of the flow table in user
space. Does a megaflow entry contain any wildcard bits?
thanks.
+
Best,
Hao
___
dis
First of all I would update to OVS 2.4 or master.
That said, the first issue you're facing might be related to a
packet leak (I suspect in the code that moves the packets to the
slow path). You could check the hit/miss counters with
`ovs-appctl dpif/show` (we added ovs-appctl dpif-netdev/pmd-stats
> On Sep 28, 2015, at 1:01 AM, 俊 赵 wrote:
>
>
> I also have a VM running openvswitch. The VM has only one port called eth0,
> and I create a bridge called s1, I add port eth0 to s1.
>
> Now I can use ssh to connect to my VM.
>
> But as soon as I set openvswitch's controller to opendaylight,
Hi all,
I have two questions.
First,
I install opendaylight lithium and install several features listed as follows
opendaylight-user@root>feature:install
odl-restconfopendaylight-user@root>feature:install
odl-openflowplugin-allopendaylight-user@root>feature:install
odl-mdsal-apidocsopendaylight
Ok thanks Justin. I’m also using a controller connected to the switch. So an
l3_learning module controller should be sufficient of the routing right?
Anyways I’ll try vxlan or GRE for tunnelling since I haven’t heard of geneve .
Thanks for your help :)
Nishanth
> On 26-Sep-2015, at 11:56 pm,
Hello Ben,
> What's your motivation?
I want to check the packet hit over a flow entry. For example I have two
flow entries as below
priority=10,dl_src=00:00:11:22:33:44, dl_dst=00:00:00:33:44:55
dl_type=0x0800 nw_proto=6 actions=2
priority=10,dl_src=00:00:11:22:33:44, dl_dst=00:00:00:33:44:55 dl
14 matches
Mail list logo
