Thanks for your quick reply.
So, is it required to "set-manager"? I did not understand the need for the
command.
Also, when I try to set the remote IP using
ovsdb-server --remote=ptcp:6640:10.107.48.192
I get the following:
2016-04-23T16:05:29Z|1|socket_util|ERR|6640:10.107.48.192: bind: C
Yes, I apologize for the mistake... I had a too quick look to one of the
last capture. Checking better the captures and the standard...
The datapath_id is included only in the OFPT_FEATURES_REPLY packet sent
from the switch to the controller during the handshake.
When you received a packet (e
Dear Ben,
I discovered the issue in visualization of port options. Indeed, "show"
command is working properly, but there is a problem in the application
of some of the options.
As a example, after applying:
> sudo ovs-ofctl mod-port br0 2 down -O OpenFlow13
> sudo ovs-ofctl mod-port br0 2 no
Hi:
This question can not be reproducibleļ¼so,it does not have a precise
trigger conditions.
At 2016-06-01 12:44:36, "Ben Pfaff" wrote:
>On Wed, Jun 01, 2016 at 11:41:00AM +0800, frankzfz wrote:
>> openvswitch version is 2.3.0,Execute the following command:
>> ovs-vsctl get bridge
>> ov
> I am running OVS 2.5v built on linux Kernel 4.3, my scenario is a 3 hops
> architecture(PUSH-SWAP-POP), but with MPLS flows I am getting an end-to -end
> throughput of 845Kb/s over a 1Gb/s NIC, while with normal IP flows I get
> 860Mb/s. I would appreciate if any one can share with me his or her
Hi
Using open vswitches v.2.5.0 with virtualization on Linux/KVM and having
random issues with cross VMs talk and wondering about increasing drops seen on
the switch it self on my hypervisor nodes when similar drops are not really
seen on any VM ports or the physical ports.
netstat
How may I
> Using open vswitches v.2.5.0 with virtualization on Linux/KVM and having
> random issues with cross VMs talk and wondering about increasing drops seen on
> the switch it self on my hypervisor nodes when similar drops are not really
> seen on any VM ports or the physical ports.
Run top on the ho
On Wed, Jun 01, 2016 at 07:03:48AM +, ojas kanhere wrote:
> So, is it required to "set-manager"? I did not understand the need for the
> command.
If you need to connect to OVSDB over TCP or SSL, then you need to use
set-manager.
> Also, when I try to set the remote IP using
>
> ovsdb-server
Na/Srilatha,
Great, I am working from three repos:
https://github.com/doonhammer/networking-sfc
https://github.com/doonhammer/networking-ovn
https://github.com/doonhammer/ovs
I had an original prototype working that used an API I created. Since then,
based on feedback from everyone I have been
On Wed, Jun 01, 2016 at 11:34:52AM +0500, Tooba Ahsen wrote:
> I need to alter the code for the select group implementation in ovs. I
> need to access the sequence numbers of the packets that come in. I
> haven't been successful so far, so can someone tell me how I should go
> about doing so?
What
> On 1. jun. 2016, at 15.42, O'Reilly, Darragh wrote:
>
>> Using open vswitches v.2.5.0 with virtualization on Linux/KVM and having
>> random issues with cross VMs talk and wondering about increasing drops seen
>> on
>> the switch it self on my hypervisor nodes when similar drops are not reall
Hi,
1. setup tc ingress policing rule on interface
2. ovs-vsctl add-port the port to ovs bridge.
3. the ingress rules will be clear by ovs.
That the default ingress_policing_rate == 0 in ovsdb clears the tc
configuration.
I think ovs should have some way to avoid overwriting user configuration.
_
On Wed, Jun 01, 2016 at 11:56:28PM +0800, Huanle Han wrote:
> Hi,
>
> 1. setup tc ingress policing rule on interface
> 2. ovs-vsctl add-port the port to ovs bridge.
> 3. the ingress rules will be clear by ovs.
>
> That the default ingress_policing_rate == 0 in ovsdb clears the tc
> configuration.
John McDowall wrote on 05/31/2016 07:57:02
PM:
> From: John McDowall
> To: Ryan Moats/Omaha/IBM@IBMUS
> Cc: Ben Pfaff , "discuss@openvswitch.org"
> , Justin Pettit ,
> "OpenStack Development Mailing List" d...@lists.openstack.org>, Russell Bryant
> Date: 05/31/2016 07:57 PM
> Subject: Re: [OVN
Please don't drop the mailing list.
OK, you're talking about TCP sequence numbers. Thanks, that helps.
The code for select groups can't efficiently access the sequence number
for a TCP packet, since that varies on a packet-by-packet basis and
would defeat caching.
On Thu, Jun 02, 2016 at 01:36:
Hi John,
Thanks your reply.
Seems you have covered everything :)
The development work can be broken down in 3 parts:
1, add ovn driver to networking-sfc
2, provide APIs in networking-ovn for networking-sfc
3, implement the sfc in ovn
So what about we take part 1 and part 2, and you take part 3?
Hi,
Suppose we have two OVS switches A and B attached to two NICs respectively.
I have a VM in between that normally bridges traffic between A and B (and
thus bridges traffic ingressing on the NICs).
Is there a way to set up a bypass mechanism, whereby switches A or B can
quickly determine that t
On Thu, Jun 02, 2016 at 12:41:32AM -0400, Lax Clarke wrote:
> Suppose we have two OVS switches A and B attached to two NICs respectively.
> I have a VM in between that normally bridges traffic between A and B (and
> thus bridges traffic ingressing on the NICs).
>
> Is there a way to set up a bypas
18 matches
Mail list logo
