On March 13, 2015 at 10:48:48 AM, Ben Pfaff (b...@nicira.com) wrote:
I'm working on figuring out the scheduling for this year's OVS
conference. I'm aiming for a pair of days in November and a location
near Palo Alto.
So far, I've heard about the following conflicts:
Nov. 12: p4.org conference (te
e minimum to test my firewall
> functionality. Can you recommend one (preferably written in python) that I
> can strip and use? Appreciate it.
You can look at OpenFaucet, if you’re OK with just OpenFlow 1.0:
http://rlenglet.github.io/openfaucet/
How many distinct UDP flows are you sending in your test?
--
Romain Lenglet
On November 24, 2014 at 7:18:07 PM, 张伟 (zhang...@126.com) wrote:
Hi Ben,
This almost drove me mad. I need your help. Just now I checked the
match_format() function. My understanding is that multipath_execute
flow, which should be the case if IPFIX export is enabled on the bridge.
To do that, please copy the output of “ovs-dpctl dump-flows” while sending some
packets.
Thanks,
—
Romain Lenglet
___
discuss mailing list
discuss@openvswitch.org
http://openvswitch.org/mailman/listinfo/discuss
commendations as to a collector please?
I’ve heard people have successfully used recent versions of nfdump to collect
IPFIX from OVS:
http://nfdump.sourceforge.net/
--
Romain Lenglet
___
discuss mailing list
discuss@openvswitch.org
http://openvswitch.org/mailman/listinfo/discuss
pipermail/dev/2014-October/047360.html
Nit in the doc for the tunnelKey element, add a space:
GENEVE(24-bit VNI) -> GENEVE (24-bit VNI)
Otherwise, those looks good to me.
Thanks,
--
Romain Lenglet
___
discuss mailing list
discuss@openvswitch.org
h
interesting in ovs-vswitchd's logs?
Also, please check with tcpdump or wireshark whether OVS actually sends IPFIX
messages out.
OVS’s IPFIX templates are quite special, so I wouldn’t be surprised if your
IPFIX collector didn’t support them.
--
Romain Lenglet
On October 29, 2014 at 12:33:
s AFAIK.
The NetFlow and IPFIX exporters are implemented quite differently in OVS.
What do you want to know?
--
Romain Lenglet
___
discuss mailing list
discuss@openvswitch.org
http://openvswitch.org/mailman/listinfo/discuss
PFIX is a beta feature in very recent versions of NFDump.
I know people are successfully parsing IPFIX flows from OVS with recent
versions.
--
Romain Lenglet
___
discuss mailing list
discuss@openvswitch.org
http://openvswitch.org/mailman/listinfo/discuss
ss data from OVS.
Commercials collectors should be usable too.
--
Romain Lenglet
___
discuss mailing list
discuss@openvswitch.org
http://openvswitch.org/mailman/listinfo/discuss
Please install the OVS kernel module that is of the same version as your
userspace OVS.
The format of userspace actions has changed in 1.11, so IPFIX sampling in OVS
1.11 will not work with an openvswitch module earlier than 1.11.
Thanks,
--
Romain Lenglet
On Oct 2, 2013, at 11:45 PM, sonia
he sampling action is failing in userspace.
This might be due to a version incompatibility between kernelspace and
userspace.
Could you please check that your OVS kernel datapath module has the same
version as your OVS userspace version?
What is the version of your OVS kernel module?
--
Romain Leng
I can reproduce, thanks.
When any traffic is sent on a bridge with both ipfix per-bridge sampling and a
controller setup, vswitchd core dumps.
I'll fix this.
--
Romain Lenglet
On Sep 3, 2013, at 2:05 PM, Glenn McGuire wrote:
> > My understanding from your message is that removin
t;From the OVS documentation: " If there are primary controllers, removing all
>of them clears the flow table."
If there are no flows, it would be logical that nothing gets sampled by IPFIX,
especially if the bridge is in "secure" fail mode.
I'll know more with the in
nw_dst, all your flows are matched.
Which flow(s) exactly did you want to match?
--
Romain Lenglet
--
Romain Lenglet
On Thursday, March 3, 2011 at 21:03, Haris Rotsos wrote:
> Hi everyone,
>
> I have a minor problem trying to implement a stats polling mechanism from an
> openflow c
X may send sensitive
information from RAM on the wire. That's not acceptable. NOX should definitely
normalize its output: zero out undefined wildcard bits, zero out wildcarded
fields, etc.
Sincerely,
--
Romain Lenglet
On Tuesday, February 22, 2011 at 03:27, kk yap wrote:
> Hi,
&g
On Wed, Jan 19, 2011 at 9:32 AM, Frank wrote:
> On 01/18/2011 04:26 PM, Frank wrote:
>
>> On 01/18/2011 04:21 PM, Romain Lenglet wrote:
>>
>>> ovs create interface name=test type=gre
>>> options:remote_ip=192.168.152.22 options:key=2
>>>
>>
to gre interface points to 192.168.152.22 under bridge xapi9.
> The first one named "3-1" was successful, the another one "test" was
> failed.
>
>
>
> On 01/18/2011 03:57 PM, Romain Lenglet wrote:
>
> Hi Frank,
> Could you send the output of "
;
> Jan 18 10:57:23|11332|bridge|INFO|created port test on bridge xapi7
> Jan 18 10:57:23|11333|bridge|WARN|could not create iface test: File exists
> Jan 18 10:57:23|11334|bridge|WARN|bridge xapi7: port test has no
> interfaces, dropping
>
>
>
>
> On 01/16/2011 04:25 PM, R
Hi Frank,
You don't create the GRE ports correctly.
The options and type are incorrect. See below.
On Fri, Jan 14, 2011 at 7:22 AM, Frank wrote:
> On 01/04/2011 02:34 PM, Jesse Gross wrote:
>
>> On Thu, Dec 30, 2010 at 6:00 PM, Frank wrote:
>>
>>> Hi guys:
>>>I just try to create two GRE t
Hi Frank,
You can't name a port the same name as your bridge, as a bridge always
explicitly has an internal interface with the same name as the bridge,
and all ports must have different names.
Try naming your GRE port differently, e.g. test2.
Regards,
--
Romain Lenglet
On Friday, Decemb
st regards,
--
Romain Lenglet
___
discuss mailing list
discuss@openvswitch.org
http://openvswitch.org/mailman/listinfo/discuss_openvswitch.org
22 matches
Mail list logo
