On Mon, 04 Mar 2019 22:07:38 +0900,
Maen Artimy wrote:
> 
> [1  <multipart/alternative (7bit)>]
> [1.1  <text/plain; UTF-8 (7bit)>]
> Diarmuid,
> I may have misunderstood your question, but the openflow does not allow the
> output port to be the same as input port.
> If you must do that then you have to use the reserved port OFPP_IN_PORT,
> which forces the output port to be the same as the input.

I don't think sending back packets is beneficial in this case.

> On Mon, Mar 4, 2019 at 5:38 AM Diarmuid O Briain <diarm...@obriain.com>
> wrote:
> 
> > Hi,
> >
> > I added the following lines to the event handler
> >
> >         if (eth.dst == eth.src):
> >             print(f'EQUAL MACs: SRC: {eth.src} == DST: {eth.dst}')
> >             return
> >
> > EQUAL MACs: SRC: 00:1e:be:17:eb:9a == DST: 00:1e:be:17:eb:9a
> > EQUAL MACs: SRC: c4:71:fe:10:fe:00 == DST: c4:71:fe:10:fe:00
> > EQUAL MACs: SRC: 00:1e:be:17:eb:9a == DST: 00:1e:be:17:eb:9a
> >
> > I see lots of equal MACs but now I get no new flows added:

Are these packets generated by your Netgear switch?

> > ~$ curl -X GET http://198.8.8.235:8080/stats/flow/176199429686713; echo
> > {"176199429686713": [{"priority": 0, "cookie": 0, "idle_timeout": 0,
> > "hard_timeout": 0, "byte_count": 51783, "duration_sec": 319,
> > "duration_nsec": 180000000, "packet_count": 648, "length": 80, "flags": 0,
> > "actions": ["OUTPUT:CONTROLLER"], "match": {}, "table_id": 0}]}
> >
> > regards,
> >
> > Diarmuid
> > --
> >
> > *Irish by birth, located in Uganda but Munster by the grace of God.*
> >
> >
> >
> >
> > On Mon, 4 Mar 2019 at 12:16, Diarmuid O Briain <diarm...@obriain.com>
> > wrote:
> >
> >> Hi,
> >>
> >> Referring to a mail I chain I started on this email group a couple of
> >> weeks ago. The support engineer in Netgear asked me for a simple
> >> description and how it could be replicated.
> >>
> >> For reference and in case anyone here is interested I attach this
> >> description and a howto replicate.
> >>
> >> Essentially the Netgear switch sends a packet for each port that has SRC
> >> MAC == DST MAC:
> >>
> >>    - IN PORT: 1
> >>    - SRC_MAC: *00:1e:be:17:eb:9a*
> >>    - DST_MAC: *00:1e:be:17:eb:9a*
> >>
> >> This causes the simple_switch_13.py to ADD FLOW with the in port and the
> >> out port the same.
> >>
> >>    - Priority: 1
> >>    - OFPMatch(oxm_fields={*'in_port': 1*, 'eth_dst':'00:1e:be:17:eb:9a'})
> >>    - [OFPActionOutput(len=16,max_len=65509,*port=1*,type=0)]
> >>    - Buffer ID: 267
> >>
> >> Would it be safe to add a loop to drop add_flow() calls where the Match
> >> port and the Action port are the same as well as the src and dst MAC
> >> addresses? Is there any condition where that would be normal behaviour?
> >>
> >> regards,
> >>
> >> Diarmuid
> >> --
> >>
> >> *Irish by birth, located in Uganda but Munster by the grace of God.*
> >>
> >>
> >> _______________________________________________
> > Ryu-devel mailing list
> > Ryu-devel@lists.sourceforge.net
> > https://lists.sourceforge.net/lists/listinfo/ryu-devel
> >
> [1.2  <text/html; UTF-8 (quoted-printable)>]
> [2  <text/plain; us-ascii (7bit)>]
> 
> [3  <text/plain; us-ascii (7bit)>]
> _______________________________________________
> Ryu-devel mailing list
> Ryu-devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/ryu-devel


_______________________________________________
Ryu-devel mailing list
Ryu-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/ryu-devel

Reply via email to