On Tue, 18 Feb 2014 09:15:01 +0800
arne_goetje wrote:
> please don't drop the integrated tests, or at least leave them as coding
No problem. I will not remove the feature if it's useful for Ryu users :)
> examples. I find them more useful than the JSON based tests, since the
> integrated test
On Tue, 18 Feb 2014 14:57:11 +0900
BYEONG-GI KIM wrote:
> I've found several OpenFlow Controller open-softwares, but I wonder whether
> Ryu is the one of such OpenFlow Contoller or not.
>
> Could you introduce Ryu in detail?
I don't know the details of other OSS OpenFlow controllers but Ryu can
On Thu, 13 Feb 2014 11:54:07 +0900
Satoshi Kobayashi wrote:
> 2014-02-13 8:23 GMT+09:00 FUJITA Tomonori :
>
>> Hi,
>>
>> On Wed, 12 Feb 2014 12:52:36 +0900
>> Satoshi Kobayashi wrote:
>>
>> > If Raw TCP is changed to WebSocket only, It do not have many merits. But
>> > when messaging is also ch
On Tue, 18 Feb 2014 10:40:47 +0900 (JST)
yamam...@valinux.co.jp (YAMAMOTO Takashi) wrote:
>> If it is only using WebSocket, I think that Gevent is good. Is using Gevent
>> and Eventlet together safe?
>
> unfortunately, no.
How about ws4py?
https://ws4py.readthedocs.org/en/latest/
Looks like no
Hi All:
Looking at the /ryu/ofproto/ofproto_v1_3.py, seems like its "unicast" mask of
"255.255.255.255" vs not?
if self._wc.ft_test(ofproto_v1_3.OFPXMT_OFB_IPV4_SRC):
if self._wc.ipv4_src_mask == UINT32_MAX:
header = ofproto_v1_3.OXM_OF_IPV4_SRC
el
Hello.
I'm a beginner who are getting started learning OpenFlow and OpenStack.
I've found several OpenFlow Controller open-softwares, but I wonder whether
Ryu is the one of such OpenFlow Contoller or not.
Could you introduce Ryu in detail?
Thanks in advance!
"The Pop PBB header action does not remove the backbone VLAN
header (B-TAG)" as described in openflow spec 1.3.3.
The format of I-TAG TCI is pcp,..,I-SID,C-DA,C-SA as in
802.1Q-2011 Figure 9-3.
---
ryu/tests/switch/of13/action/27_POP_PBB.json |3 +++
1 files changed, 3 insertions(+), 0 deletio
> Hi,
>
> If it is only using WebSocket, I think that Gevent is good. Is using Gevent
> and Eventlet together safe?
unfortunately, no.
YAMAMOTO Takashi
>
> 2014-02-14 22:40 GMT+09:00 Jeremias Blendin <
> jblen...@kom.e-technik.tu-darmstadt.de>:
>
>> Hi,
>>
>> I have used gevent websocket, whi
Hi,
If it is only using WebSocket, I think that Gevent is good. Is using Gevent
and Eventlet together safe?
2014-02-14 22:40 GMT+09:00 Jeremias Blendin <
jblen...@kom.e-technik.tu-darmstadt.de>:
> Hi,
>
> I have used gevent websocket, which integrated well with pywsgi, which is
> (or was?) used
Dear Fujita-san,
please don't drop the integrated tests, or at least leave them as coding
examples. I find them more useful than the JSON based tests, since the
integrated tests allow me to perform certain tests only on certain
conditions. I.e. I query the switch capabilities first and then dec
Hi Fujita,
Thank you so much for the response.
What I am trying to achieve is incrementally pass flow with the combination
of matches. For Example, I first pass a flow with match on ethertype, next
with match on ethertype and source MAC, then with match on ethertype,
source MAC and destination MA
On Mon, 17 Feb 2014 17:17:46 +0900 (JST)
yamam...@valinux.co.jp (YAMAMOTO Takashi) wrote:
> > I believe that it makes sense to treat a missing mask
> > as not being an all-zero bit mask.
> >
> > The motivation for this change is to resolve the following problem
> > that I observed when running Op
On Mon, 17 Feb 2014 12:49:08 +0900
YAMAMOTO Takashi wrote:
> suggested by some openstack folks.
> see https://review.openstack.org/#/c/71791
>
> Signed-off-by: YAMAMOTO Takashi
> ---
> bin/neutron-ofagent-agent | 19 +++
> bin/neutron-ofswitch-agent| 19
On Mon, 17 Feb 2014 12:21:34 +0900 (JST)
yamam...@valinux.co.jp (YAMAMOTO Takashi) wrote:
> > On Sat, 15 Feb 2014 21:20:23 +
> > Morgan Yang wrote:
> >
> >> I'm assuming this is a typo?
> >
> > Yeah, looks like so. How about the following?
>
> looks good to me.
>
> Reviewed-by: YAMAMOTO T
On Mon, 17 Feb 2014 15:28:18 +0900
Simon Horman wrote:
> On Mon, Feb 17, 2014 at 12:55:08PM +0900, YAMAMOTO Takashi wrote:
> > Signed-off-by: YAMAMOTO Takashi
>
> Reviewed-by: Simon Horman
Applied, thanks.
--
Android
On Mon, 17 Feb 2014 12:17:19 +0900 (JST)
yamam...@valinux.co.jp (YAMAMOTO Takashi) wrote:
> > They are necessary for switches that don't support the feature of
> > modifying 'running' configuration directly; instead, needs to modify
> > 'candidate' configuration and then commit (or discard).
> >
On Mon, Feb 17, 2014 at 05:17:46PM +0900, YAMAMOTO Takashi wrote:
> > I believe that it makes sense to treat a missing mask
> > as not being an all-zero bit mask.
> >
> > The motivation for this change is to resolve the following problem
> > that I observed when running Open vSwtich's make check-r
> I believe that it makes sense to treat a missing mask
> as not being an all-zero bit mask.
>
> The motivation for this change is to resolve the following problem
> that I observed when running Open vSwtich's make check-ryu target.
the real problem is that ovs check-ryu creates two bridges to co
18 matches
Mail list logo
