At Wed, 20 Jul 2016 13:36:41 +0900,
Shinpei Muraoka wrote:
>
> This patch will revert the API of NXAction for the following report.
> https://bugs.launchpad.net/dragonflow/+bug/1599367
> revert commit id: a01f8f0aa03cd85df2fc07e7217c2a79e7f0
You meant commit 3c3896fd77f274ce953dcb78795aa5cc4f
Hi Iwase,
Thanks for your response.
I agree, RYU does close down applications cleanly with the call to
app_mgr.close() after a KeyboardInterrupt. I have hooks in my application to
close() cleanly in this event.
However I was hoping that the console output (i.e. the Traceback.
KeyboardInte
Suggest just silently swallowing the KeyboardInterrupt (with a "pass"), since
the finally block will cover the close().
Best,
Victor
--
Victor J. Orlikowski <> vjo@[cs.]duke.edu
> On Jul 20, 2016, at 10:27 AM, "alan_ford...@keysight.com"
> wrote:
>
>except KeyboardInterrupt:
>app_
Also - should be obvious - but include a comment on the pass, explaining that
you're "papering over" the fact that eventlet doesn't properly wrap the
KeyboardInterrupt inside select()/poll()/epoll().
Another suggestion: instead of a pass, a log message, at debug level (or
whatever desired defau
Your suggestion of a log message is probably the best approach.
Thanks,
Alan.
-Original Message-
From: Victor Orlikowski [mailto:v...@duke.edu]
Sent: Wednesday, July 20, 2016 11:36 AM
To: FORDYCE,ALAN (K-Scotland,ex1)
Cc: ryu-devel@lists.sourceforge.net
Subject: Re: [Ryu-devel] Keyboard
Hi Iwase,
Taking on board Victor's comments I made the following changes to my manager.py
file...
LOG = logging.getLogger('ryu.cmd.manager')
if CONF.enable_debugger:
msg = 'debugging is available (--enable-debugger option is turned on)'
LOG.info(msg)
else:
hu
Hi Ryu development team,
I'm doing a small project using Ryu applications and encountering on small
issue. I want to run two applications in parallel as 2 web servers on 2
different wsapi-ports but with only one controller:
ryu-manager ryu.app. ryu.app
However, I could not tell ryu-manager to d
Hi Iwase,
After including module openvswitch, ryu-manager call is able to move
further. But it is still not able to complete things.
instantiating app ryu.controller.ofp_handler of OFPHandler
BRICK SimpleSwitch13
CONSUMES EventOFPSwitchFeatures
CONSUMES EventOFPPacketIn
BRICK ofp_event
PROV
Thanks, Iwase. It worked perfectly.
I think that BGP support is a strategy to facilitate a communication
between SDN networks and non-SDN networks. I mean, there's no reason to a
controller to implement a standard routing protocol like BGP in a
SDN-enabled environment. Am I right?
Helcio
Hi,
On 2016年07月21日 02:41, Deepak Jose wrote:
> Hi Iwase,
>
> After including module openvswitch, ryu-manager call is able to move further.
> But it is still not able to complete things.
>
> instantiating app ryu.controller.ofp_handler of OFPHandler
> BRICK SimpleSwitch13
> CONSUMES EventOFPSwi
Hi Helcio,
BGP is also useful on SDN networks, I think.
For example, BGP can exchange the network information which managed one SDN
controller to other SDN controllers, so SDN controllers can cooperate each
other in the controllers cluster or so.
Therefore, it is reasonable to support BGP service
Hi Iwamoto-San
> You meant commit 3c3896fd77f274ce953dcb78795aa5cc4fde6c65?
Sorry it was my mistake.
Modify commit ID to 3c3896fd77f274ce953dcb78795aa5cc4fde6c65.
> I think it's better to separate this into the pure revert patch and
> the other changes in order to make git history more read
Hi Alan and Victor,
Thank you very much for your suggestions and patch!
But, I am just one of the contributers for Ryu.
Then, would you review my attached patch?
If this patch looks good, this might be adopted through Ryu-devel.
FYI, with this patch, ryu-manager will outputs messages as follow
On Thu, 21 Jul 2016 10:19:07 +0900
Shinpei Muraoka wrote:
> > I think it's better to separate this into the pure revert patch and
> > the other changes in order to make git history more readable.
>
> Separate this into the revert patch and the other changes.
> However, revert patch has an impa
Looks good to me, but I have not run it. ;)
A couple of questions/comments:
1) I presume you have verified that all uses of "LOG" have been changed to
"logger"?
Given the scoping, It appears you should have gotten them all.
2) I would re-write the commit message thus:
Currently, ryu-manager emit
[image: Inline image 1]
As per the ryubook doc, I thought we should get above pasted messages. Is
that wrong?
On Thu, Jul 21, 2016 at 6:01 AM, Iwase Yusuke
wrote:
> Hi,
>
>
> On 2016年07月21日 02:41, Deepak Jose wrote:
>
>> Hi Iwase,
>>
>> After including module openvswitch, ryu-manager call is abl
Hi Victor,
Thank you for reviewing my patch.
On 2016年07月21日 11:32, Victor Orlikowski wrote:
Looks good to me, but I have not run it. ;)
A couple of questions/comments:
1) I presume you have verified that all uses of "LOG" have been changed to
"logger"?
Given the scoping, It appears you shoul
Alan:
It seems that you have to install a ZK server first.
But, these code have not merged into upstream for some reasons.
Iwase:
BTW, Are there any roadmap of RYU to support HA instead of
Implementing it in user app?
On Wed, Jul 20, 2016 at 1:51 PM, Alan Wang wrote:
> Hi All,
>
> Thanks for you
Hi,
Yes, you are right.
But, there seems to have been a little misunderstanding.
On Ryu-Book, the tree dots("...") means that some messages or words are just
omitted.
e.g.)
If the full messages are:
---
connected socket: address:('127.0.0.1', 32776)
hello ev
move onto config mode
Hi,
On 2016年07月21日 13:11, Shuoling Deng wrote:
> Alan:
> It seems that you have to install a ZK server first.
> But, these code have not merged into upstream for some reasons.
>
> Iwase:
> BTW, Are there any roadmap of RYU to support HA instead of
> Implementing it in user app?
Sorry, I don't rea
For the report from Wisam Maala (*1), this patch fixes rest_qos.py
to enable to parse the flow which does not contain any QoS action.
(*1)
https://sourceforge.net/p/ryu/mailman/ryu-devel/thread/1005235784.4054991.1468592243418.JavaMail.yahoo%40mail.yahoo.com/#msg35222959
IWASE Yusuke (1):
rest
Currently, rest_qos.py might fail to parse the actions field in
the QoS rules when the actions field does not have any actions
for the QoS (SET_FIELD, METER, SET_QUEUE).
This patch fixes this problem.
Signed-off-by: IWASE Yusuke
---
ryu/app/rest_qos.py | 14 +++---
1 file changed, 7 ins
Hi,
Thanks for responding. The patch that you suggested totally fix the
problem.
Best Regards,
Hong Panha
On Wed, Jul 20, 2016 at 2:54 PM, Iwase Yusuke
wrote:
> Hi,
>
> The cause of this problem seems that the difference of OFPT_PORT_STATUS
> message
> behavior between Open vSwitch(OVS) and La
So I'm good to continue my exercise. Right?
Thanks for the help Iwase.
On Thu, Jul 21, 2016 at 9:42 AM, Iwase Yusuke
wrote:
> Hi,
>
> Yes, you are right.
>
> But, there seems to have been a little misunderstanding.
> On Ryu-Book, the tree dots("...") means that some messages or words are
> just
No - I think you were right the first time; go ahead and follow the naming
convention. ;)
So long as you made sure to check that LOG isn't referenced anywhere else,
all's well.
Best,
Victor
--
Victor J. Orlikowski <> vjo@[cs.]duke.edu
> On Jul 21, 2016, at 4:59 AM, Iwase Yusuke wrote:
>
> FY
Hi Fujita-San
> The patch can't be cleanly reverted. So let's forget this point.
I got it.
I will include it in the patch.
> Keeping the jsondict outputs is nice but it's optional for me.
I think the simple source is good.
I modify the format to be set the zone_src a string.
If you want to s
Hi,
I think it's yes.
If get another problem during the training on Ryu-Book,
feel free to contact us via Ryu-devel mailing list.
Thanks,
Iwase
On 2016年07月21日 15:42, Deepak Jose wrote:
> So I'm good to continue my exercise. Right?
> Thanks for the help Iwase.
>
> On Thu, Jul 21, 2016 at 9:42 AM,
27 matches
Mail list logo
