On 22 February 2017 at 09:53, Sanjeev Maheve
wrote:
> Hello Teams,
>I am using openvswitch for connecting Docker containers
> on multiple hosts using OVS (overlay networks - Geneve encapsulation). I
> have a requirement to connect Docker containers
Please send your mails in plain text.
Did you run the "req" command in the same directory. You need to do
something like this:
https://github.com/openvswitch/ovn-kubernetes/blob/master/vagrant/provisioning/setup-master.sh#L66
On 19 February 2017 at 23:29, 冷月无声 <645363...@qq.com> wrote:
> Dear
On 22 February 2017 at 09:53, Sanjeev Maheve
wrote:
> Hello Teams,
>I am using openvswitch for connecting Docker containers
> on multiple hosts using OVS (overlay networks - Geneve encapsulation). I
> have a requirement to connect Docker containers
On 6 December 2016 at 20:01, Joo Yong-Seok wrote:
> Hello,
>
> Is there any good example for openvswitch flow/rules for captive-portal?
> Which means,
>
> - We should perform DNAT (with captive-portal web server IP) for outbound
> HTTP traffic
> - When responses are back,
* Firstly, getting 10 Gbps for stock vxlan is surprising. Are you using
vxlan hardware offload capable NICs?
* Such drastic performance difference is not expected.
* Give a very simple example of how you are using flow based VXLAN. i.e the
actual flow.
* What is the kernel module that you are
ct_lb is tricky. I guess, the default should be to just pick the first
option.
On 30 November 2016 at 08:24, Ben Pfaff wrote:
> One initial, trivial, step might be to just have every ct_next respond
> that the flow is established. Then at least it would be possible to see
> how
aware of any. One needs to be comfortable with OVS and ovs-ofctl
to understand most of these. You can also read 'man ovs-ofctl' to read
about more details around 'ct' action.
>
> Best regards
>
> - yongseok
>
> On Wed, Jan 4, 2017 at 8:42 AM, Guru Shetty <g...@ovn.org> wr
On 21 March 2017 at 09:01, 蘇于倫 wrote:
> Hi everyone,
>
> I'd like to confirm which steps should I do to completely remove current
> OVS if I want to re-install OVS with modified C code in git repo ?
>
> Currently what my reinstall_ovs.sh does for old OVS removal
;
> It shows only one instance.
> Does it make sense if I have re-installed OVS?
>
> Ethan
>
>
>
> 2017年3月22日 01:04,"Guru Shetty" <g...@ovn.org>寫道:
>
>
>
> On 21 March 2017 at 09:01, 蘇于倫 <yulunsu.eed...@g2.nctu.edu.tw> wrote:
>
>> Hi e
On 23 March 2017 at 22:36, Shivaram Mysore
wrote:
> One more data point:
>
> I *CNTRL-C* the command ovs-vsctl add-port
>
> ^C2017-03-24T05:25:30Z|2|fatal_signal|WARN|terminating with signal 2
> (Interrupt)
>
>
> # *ovs-vsctl show*
>
Did you have a look at the couple we have here?
http://docs.openvswitch.org/en/latest/howto/docker/
On 12 April 2017 at 09:07, Hui Kang wrote:
>
> Hi All,
> Please checkout the libnetwork plugin for OVN at
> https://github.com/huikang/libnetwork-ovn-plugin.
> Comment and PR
It looks like you already figured it out.
On 19 July 2017 at 07:28, Zekeriya Akgül wrote:
> Hi,
> I am student on Necmettin Erbakan Universty in Turkey.And working on
> communication between dpdk powered openvswitch and docker.And some issues
> occured about communication
On 24 April 2017 at 11:31, Hexin Wang wrote:
> Hi,
>
> Is there any knob in OVN to control when floating IP will be applied in
> the distributed NAT? Specifically:
>
>1. If the destination IP is part of some private layer3 domain, the
>usual private IP is used to
P is used for south-north default route to internet traffic.
> E.g. VM originated internet traffic: SNAT is applied to change source IP
> to floating IP. UNSNAT is applied to change destination IP back to private
> IP.
>
> Yes. There are examples in tests/system-traffic.at
> Tha
o please try again to frame your question without depending on previous
mails for context. With a clear example.
>
> Thanks.
>
> Hexin
>
> From: <ovs-discuss-boun...@openvswitch.org> on behalf of Hexin Wang
> Date: Monday, April 24, 2017 at 2:57 PM
> To: Guru Shetty
>
On 9 June 2017 at 20:28, Arunkumar Rg wrote:
> HI All,
>
> Forgot to mention, my query pertains to hardware_vtep schema based DB.
>
hardware_vtep schema does not have the SSL table. ovsdb-server does take
multiple databases as arguments. So one idea (I haven't tried it)
tcp:3332:192.168.1.2 -- blah
> --
> *From:* ovs-discuss-boun...@openvswitch.org <ovs-discuss-bounces@
> openvswitch.org> on behalf of Omar Ramadan <orama...@fb.com>
> *Sent:* Thursday, October 12, 2017 4:47:10 PM
>
> *To:* Guru Shetty
> *Cc:* ovs-discuss@openvswitch.or
On 12 October 2017 at 10:29, Omar Ramadan wrote:
> Hi list,
>
>
> I am using OVS 2.7.90 with Debian 8.7 and want to configure the switch to
> be loaded on system initialization. I have installed the service
> "openvswitch-switch" and added the following in
;
> Do I need to register new a ovs_tunnel_type somewhere for it to work? Any
> way to increase logging here?
>
You can add a "set -x" to the top of
/usr/share/openvswitch/scripts/ifupdown.sh
And then run ifup and ifdown.
>
> Best,
> Omar
>
> --
On 15 November 2017 at 01:17, Dávid Patrik Pintér <
david.patrik.pin...@ericsson.com> wrote:
> Hi All,
>
> I have a question about the OVS-Bugtool plugins:
> What's the actual benefit of using bash scripts for calling single line
> shell commands:
>
On 14 November 2017 at 14:40, Sébastien Bernard wrote:
> Hello,
>
> I'm looking for some pieces of advise to use a network based on
> openvswitch with kubernetes.
>
> I've tried to follow the following document https://github.com/openvswitch
> /ovn-kubernetes, with some
On 16 November 2017 at 01:56, Sébastien Bernard wrote:
> Ok,
>
> I got to reproduce the error I had yesterday.
>
> Here's the path :
>
> 1- one vm with centos 7
>
> 2- install kubeadm v1.8.3
>
> 3- kubeadm init
>
> 4- install openvswitch (v2.8.1)
>
> 5- follow the
On 16 November 2017 at 09:26, Sébastien Bernard <sbern...@nerim.net> wrote:
> On 16/11/2017 13:08, Guru Shetty wrote:
>
>
>
> On 16 November 2017 at 01:56, Sébastien Bernard <sbern...@nerim.net>
> wrote:
>
>> Ok,
>>
>> I got to reprodu
On 15 November 2017 at 05:02, Sébastien Bernard <sbern...@nerim.net> wrote:
> On 15/11/2017 01:22, Guru Shetty wrote:
>
>
>
> On 14 November 2017 at 14:40, Sébastien Bernard <sbern...@nerim.net>
> wrote:
>
>> Hello,
>>
>> I'm looki
On 12 November 2017 at 22:43, Hui Xiang wrote:
> Does ovs linux dapath NAT work with linux kernel 4.4.70 version?
>
If you use the kernel module that comes with OVS repo, it will work. If you
use the kernel module that comes by default with linux kernel, it won't.
You can
On 2 November 2017 at 13:07, kevin parrikar
wrote:
> Hello All,
> I am running OVS 2.6.1 on Ubuntu 14.04 kernel 4.4.0-87-generic with
> Openstack Mitaka release with OVS firewall driver(contrack )
>
>
> MTU is set to 9000 on both the physical nics and icmp is success
On 1 November 2017 at 03:05, Wei Li wrote:
> Hello everyone
>
> In https://github.com/openvswitch/ovs/blob/master/ovn/lib/
> actions.c?utf8=%E2%9C%93#L833
>
> } else if (snat && ep->is_gateway_router) {
> /* For performance reasons, we try to prevent additional
>
In tests/ovn.at, search for "ipam".
On 17 June 2018 at 14:27, Paul Greenberg wrote:
> All,
>
> I want to get an IP address from OVN without using DHCP. I did not find a
> command line option to do so.
>
> How could one do it through OVSDB queries? Did anyone attempt it?
>
> Best Regards,
> Paul
gt; Best Regards,
>
> Wei-Yu Chen
> Wireless Internet Laboratory
> Department of Computer Science
> National Chiao Tung University
>
> On 30 April 2018 at 11:49:29 PM, Guru Shetty (g...@ovn.org) wrote:
>
>
>
> On 26 April 2018 at 06:41, Wei-Yu Chen <aweimeow...
I was able to reproduce it. I will work with Ben to get this fixed.
On 26 July 2018 at 23:14, Girish Moodalbail wrote:
> Hello Ben,
>
> Sorry, got distracted with something else at work. I am still able to
> reproduce the issue, and this is what I have and what I did
> (if you need the core,
How did you configure OVS?
You should use:
./configure --prefix=/usr --localstatedir=/var --sysconfdir=/etc
--enable-ssl
On 25 July 2018 at 23:17, Vikas Kumar wrote:
>
> hi Team,
> when i am trying to restart the vswitch daemon i am getting the below erro:
>
>
Hello All,
Looks like with OVS 2.10, we cannot create more than 2 userspace geneve
tunnels.
You get an error that says: could not add network device geneve23 to
ofproto (Device or resource busy)
e.g:
vagrant@k8smaster:~$ sudo ovs-vsctl show
86d11177-d982-48f9-a08e-c9f6f6587215
Bridge "br0"
I should clarify that I am seeing this in kernel datapath. I realized that
my usage of "userspace geneve tunnels" can be misleading.
On Mon, 20 Aug 2018 at 21:22, Guru Shetty wrote:
> Hello All,
> Looks like with OVS 2.10, we cannot create more than 2 userspace geneve
>
A 'make install' does not install the startup script. You can copy it over
manually. E.g : cp debian/openvswitch-switch.init
/etc/init.d/openvswitch-switch
On 23 July 2018 at 01:23, Vikas Kumar wrote:
> hi Team,
> i was trying to restart the openvswitch daemon using the below script:
>
>
>
> So, if I build OVS from source, am I going to have the same problem? It
> seems that everything in Squeeze is systemd? I guess I could try and
> see. Will the init.d script work?
>
>
You said it is OVS 2.6.2. First thing to do is to simply modify your
/etc/init.d/openvswitch-switch to
Since this is the same host, your 25 Gbps physical network does not come to
picture. Neither does VXLAN as both the VMs are in the same host. I would
suggest trying the following experiments.
1. Multiple sessions of netperf/iperf.
2. Sometimes, your VM's virtual NIC is the bottleneck. So if you
rovide /etc/init.d/openvswitc-switch script).
>
> brian
>
> On Tue, Jul 03, 2018 at 10:39:03AM -0700, Guru Shetty wrote:
> >On 1 July 2018 at 22:37, Brian E. Lavender <[1]br...@brie.com> wrote:
> >
> > Turns out OVS got stuck in the systemd cycle.
On 1 July 2018 at 22:37, Brian E. Lavender wrote:
> Turns out OVS got stuck in the systemd cycle. After waiting 5 minutes,
> it came up.
>
> You are likely using packages from Canonical and not from this repo?
> brian
>
> On Sun, Jul 01, 2018 at 09:06:48AM -0700, Brian E. Lavender wrote:
> > I
On 9 March 2018 at 11:19, Ben Pfaff <b...@ovn.org> wrote:
> On Fri, Mar 02, 2018 at 09:40:07AM -0800, Guru Shetty wrote:
> > On 1 March 2018 at 21:09, Anil Venkata <anilvenk...@redhat.com> wrote:
> >
> > >
> > >
> > > On Fri, Ma
On 1 March 2018 at 21:09, Anil Venkata <anilvenk...@redhat.com> wrote:
>
>
> On Fri, Mar 2, 2018 at 7:23 AM, Guru Shetty <g...@ovn.org> wrote:
>
>>
>>
>> On 27 February 2018 at 03:13, Anil Venkata <anilvenk...@redhat.com>
>> wrote:
>&
On 27 February 2018 at 03:13, Anil Venkata wrote:
> For example, I have a 10.1.0.0/24 network and a load balancer is added to
> it with 10.1.0.10 as VIP and 10.1.0.2(MAC 50:54:00:00:00:01), 10.1.0.3(MAC
> 50:54:00:00:00:02) as members.
> ovn-nbctl create load_balancer
On 28 June 2018 at 11:58, Stephen V. Nesbitt
wrote:
> Noob alert! I am looking for a recipe that will configure an OVS defined
> network at boot on Ubuntu bionic.
>
> My ovs network comprises the following:
>
> * switch0
>
> * enp3s0 - physical NIC defined as a port tagged with 100 and marked as
On Wed, 26 Sep 2018 at 12:59, Jean-Philippe Méthot via discuss <
ovs-discuss@openvswitch.org> wrote:
> Hi,
>
> I’ve been using openvswitch for my networking backend on openstack for
> several years now. Lately, as our network has grown, we’ve started noticing
> some intermittent packet drop
very high.
>
>
> Our network nodes CPU are literally sleeping. Is openvswitch single-thread
> or multi-thread though? If ovs overloaded a single thread, it’s possible I
> may have missed it.
>
> Jean-Philippe Méthot
> Openstack system administrator
> Administrateur système Open
a_cert --user
> openvswitch:hugetlbfs --no-chdir
> --log-file=/var/log/openvswitch/ovsdb-server.log
> --pidfile=/var/run/openvswitch/ovsdb-server.pid --detach
> openvsw+ 1013 1 11 19:45 ?00:17:46 ovs-vswitchd
> unix:/var/run/openvswitch/db.sock -vconsole:emer -vsyslog:err -
On Tue, 13 Nov 2018 at 21:53, Sam wrote:
> And why OVS take high CPU cost?
>
My simplistic guess is that you have created a loop in your network with
OVS. Or your SDN flows are inefficient. For a simple setup, there should
really be not much difference with CPU between linux bridge and OVS.
On Sun, 7 Apr 2019 at 11:39, Tu Nguyen Phuoc
wrote:
> Hi OVS Community.
> I have 1 Physical host which had docker and ovs installed.
> I want to set up 2 ovs bridges (ovsbr1 and ovsbr2), each bridge has 1
> container connects to (h1 and h2), here is my set up:
> On Physical host
> ***
How do you define "poor" performance? If it is awfully low, it is likely
that your [ outer MTU < (inner MTU + vxlan header)]. You can reduce your
inner MTU and see the difference.
On Mon, 24 Jun 2019 at 06:34, Heim, Dennis wrote:
> I am running Ubuntu on virtualized hosts, along with VXLAN.
What were the errors encountered?
On Mon, 12 Aug 2019 at 21:20, V Sai Surya Laxman Rao Bellala <
laxmanraobell...@gmail.com> wrote:
> Hello all,
>
> Does anyone have a standard procedure for installing the ovs-2.11.1 in
> ubuntu 16.04 ?
> I am encountering so many problemsErrors were encountered
Did you install OVS with "make install"? If so, you will have to provide
python path for OVS python libraries. If you install it via packages, you
will need "python-openvswitch" package.
On Wed, 21 Aug 2019 at 14:46, Massimiliano Bavelloni
wrote:
> Hello,
>
> after compiling (gcc)
I had created a openvswitch repo in docker as a placeholder. Happy to
provide it to whoever the admin is.
On Thu, 7 Nov 2019 at 13:15, aginwala wrote:
> Hi All:
>
> As discussed in the meeting today, we all agreed that it will be a good
> idea to push docker images for each new ovs/ovn stable
On Fri, 8 Nov 2019 at 09:53, Guru Shetty wrote:
> I had created a openvswitch repo in docker as a placeholder. Happy to
> provide it to whoever the admin is.
>
i.e. You can use the keyword "openvswitch". For e.g., right now, it has one
stale image.
docker run -d --n
0:17 AM aginwala wrote:
>
>> Thanks Guru:
>>
>> Sounds good. Can you please grant user aginwala as admin? I can create
>> two repos ovs and ovn under openvswitch org and can push new stable release
>> versions there.
>>
>> On Fri, Nov 8, 2019 at 10:0
t; container.
>
It is not clear to me. Is DKMS enabled here? Or is it that
openvswitch/ovs:2.12.0_debian_4.15.0-66-generic will only work on
kernel 4.15.0-66-generic?
>
> On Fri, Nov 8, 2019 at 2:37 PM Guru Shetty wrote:
>
>>
>>
>> On Fri, 8 Nov 2019 at 14:18, aginwala wrote:
&g
You cannot accomplish this via nmcli as it needs OVS processes to tear down
OVS artifcats.
You clearly mention that you don't want to use OVS commands - but if your
goal is to not run OVS processes, you can potentially try your luck with
"ovs-dpctl del-dp system@ovs-system" when OVS processes are
On Mon, 11 Nov 2019 at 10:08, aginwala wrote:
>
>
> On Mon, Nov 11, 2019 at 9:00 AM Guru Shetty wrote:
>
>>
>>
>> On Fri, 8 Nov 2019 at 14:41, aginwala wrote:
>>
>>> openvswitch.ko ships default with newer kernel but if we want to use say
>&g
It may have come from this commit instead:
commit 9b5422a98f817b9f2a1f8224cab7e1a8d0bbba1f
Author: Ilya Maximets
Date: Wed Dec 16 15:32:21 2015 +0300
ovs-lib: Try to call exit before killing.
While killing OVS may not free all allocated resources.
Previously we used to SIGTERM
I have recently built OVS DPDK rpms. For e.g., yesterday, I built OVS 2.13
with DPDK 19.11
The rhel/openvswitch-fedora.spec.in should have everything you need.
You need to first build dpdk devel and dpdk rpms from DPDK. And then
install those rpms. And then:
./boot.sh
./configure
make
in/env python$|#!/usr/bin/python3|' {} +
+
%files
%dir %{_datadir}/dpdk
%{_datadir}/dpdk/usertools
@@ -81,11 +75,9 @@
%{_includedir}/dpdk
%{_datadir}/dpdk/mk
%{_datadir}/dpdk/buildtools
-%{_datadir}/dpdk/%{target}
+%{_datadir}/dpdk/arm64-%{machine}-linux-gcc
%{_datadir}/dpdk/examples
-%files d
59 matches
Mail list logo