The kernel support for Geneve is already there.
https://github.com/mchalla/ovs/tree/geneve
It was posted as an RFC a while back. I have not gotten a chance to
work on it since. I will be sending out v2 in couple of weeks.
Thanks.
On Mon, Mar 9, 2015 at 1:53 AM, Neelakantam Gaddam
wrote:
> Hi A
Like Jesse pointed out the numbers should be much better with offload. Here
are the numbers I see. My offload card is an eval card and is messing up
some packets. I am going to try it with latest firmware and update the
slides.
http://www.slideshare.net/MadhusudanChalla/ovs-perf
Thanks.
On Mon,
On Sat, Nov 8, 2014 at 11:59 AM, Jesse Gross wrote:
> On Sat, Nov 8, 2014 at 11:41 AM, Madhu Challa
> wrote:
> > On Fri, Nov 7, 2014 at 5:14 PM, Jesse Gross wrote:
> >> On Fri, Nov 7, 2014 at 3:06 PM, Madhu Challa
> wrote:
> >>> On Fri, Nov 7, 2014 at 2:
On Fri, Nov 7, 2014 at 5:14 PM, Jesse Gross wrote:
> On Fri, Nov 7, 2014 at 3:06 PM, Madhu Challa wrote:
>> On Fri, Nov 7, 2014 at 2:42 PM, Jesse Gross wrote:
>>> On Thu, Nov 6, 2014 at 3:43 PM, Madhu Challa
>>> wrote:
>>>> On Thu, Nov 6, 2014 at 2:37
On Fri, Nov 7, 2014 at 8:13 PM, Jesse Gross wrote:
> On Fri, Nov 7, 2014 at 3:12 PM, Madhu Challa wrote:
>> I was referring to the bit mask present in the spec:
>>
>> A.2.3.5 Flow Match Field Masking
>> When oxm_hasmask is 1, the OXM TLV contains a bitmask and i
, 2014 at 2:43 PM, Jesse Gross wrote:
> I'm not sure I understand what you are referring to. Can you elaborate?
>
> On Thu, Nov 6, 2014 at 3:58 PM, Madhu Challa wrote:
>> Since we are on this topic I had one more question. This could still
>> be an issue if we want to be
On Fri, Nov 7, 2014 at 2:42 PM, Jesse Gross wrote:
> On Thu, Nov 6, 2014 at 3:43 PM, Madhu Challa wrote:
>> On Thu, Nov 6, 2014 at 2:37 PM, Jesse Gross wrote:
>>>
>>> On Thu, Nov 6, 2014 at 10:08 AM, Madhu Challa
>>> wrote:
>>> > Jesse,
>
, Madhu Challa wrote:
> On Thu, Nov 6, 2014 at 2:37 PM, Jesse Gross wrote:
>>
>> On Thu, Nov 6, 2014 at 10:08 AM, Madhu Challa
>> wrote:
>> > Jesse,
>> >
>> > Thanks for sharing your thoughts on this.
>> >
>> > On Thu, Nov 6, 2014
On Thu, Nov 6, 2014 at 3:09 PM, Ben Pfaff wrote:
> On Thu, Nov 06, 2014 at 03:02:45PM -0800, Jesse Gross wrote:
>> On Thu, Nov 6, 2014 at 2:49 PM, Ben Pfaff wrote:
>> > Have you considered allowing database configuration to specify what kind
>> > of attributes are expected and how to parse them?
On Thu, Nov 6, 2014 at 2:49 PM, Ben Pfaff wrote:
> On Thu, Nov 06, 2014 at 02:37:01PM -0800, Jesse Gross wrote:
>> I'm curious how you ended up laying this out. The OpenFlow spec says
>> that the extra space should be used as an vendor ID in the form of an
>> OUI. How did you reconcile this?
>
> Y
On Thu, Nov 6, 2014 at 2:37 PM, Jesse Gross wrote:
>
> On Thu, Nov 6, 2014 at 10:08 AM, Madhu Challa
> wrote:
> > Jesse,
> >
> > Thanks for sharing your thoughts on this.
> >
> > On Thu, Nov 6, 2014 at 7:47 AM, Jesse Gross wrote:
> >>
&g
On Thu, Nov 6, 2014 at 10:40 AM, Ben Pfaff wrote:
> On Thu, Nov 06, 2014 at 10:08:00AM -0800, Madhu Challa wrote:
> > > One issue that comes up when doing this is that the TLVs in both
> > > Geneve and OXM are exactly the same size so mapping them directly
> > >
Agreed and looks good.
Thanks.
On Thu, Nov 6, 2014 at 10:57 AM, Ben Pfaff wrote:
> On Wed, Nov 05, 2014 at 06:34:04PM -0800, Madhu Challa wrote:
> > The other issue was in the same file. I am glad make check catches these
> > issues and hope there are none in other files. I a
Jesse,
Thanks for sharing your thoughts on this.
On Thu, Nov 6, 2014 at 7:47 AM, Jesse Gross wrote:
> On Wed, Nov 5, 2014 at 10:03 AM, Madhu Challa
> wrote:
> > Thanks Ben. I will debug and get back to you. I will check with Jesse in
> the
> > upcoming ovs conference if
The other issue was in the same file. I am glad make check catches these
issues and hope there are none in other files. I attached a patch that
should make it simpler to extend the union in the future.
Thanks.
On Wed, Nov 5, 2014 at 11:54 AM, Madhu Challa
wrote:
> This test was failing beca
mf_value rather than a hard coded value. I will be happy to provide a
patch after I debug the remaining failures.
Thanks.
On Wed, Nov 5, 2014 at 10:03 AM, Madhu Challa
wrote:
> Thanks Ben. I will debug and get back to you. I will check with Jesse in
> the upcoming ovs conference if he has
Thanks Ben. I will debug and get back to you. I will check with Jesse in
the upcoming ovs conference if he has other thoughts on implementing this.
Thanks.
On Wed, Nov 5, 2014 at 8:17 AM, Ben Pfaff wrote:
> On Tue, Nov 04, 2014 at 03:10:59PM -0800, Madhu Challa wrote:
> > I wa
I was playing increasing the size of these structures to be able to read
Geneve metadata. Currently they are limited to 16 bytes. I noticed that
some of the test cases fail when I do this. I guess I can read the oxms
directly without using these structures. But figured its less error prone
if done
When tracing the tunnel / vxlan path I figured this function gets used
there, after the vxlan decap and re-injecting the packet into the ip stack.
May be thats the only use case.
Thanks.
On Fri, Jun 6, 2014 at 6:51 PM, Madhu Challa
wrote:
> I have the following config
>
> b74843d3-
I was going through some of the OpenFlow 1.4 items.
OpenFlow 1.4 section 7.4.4 ``Controller Role Status Message''
defines a new message sent by a switch to notify the controller that
its role (whether it is a master or a slave) has changed. OVS should
implement this.
"When a controller has its
realistic.
Thanks.
On Wed, May 28, 2014 at 5:27 PM, Jesse Gross wrote:
> On Mon, May 26, 2014 at 5:03 PM, Madhu Challa
> wrote:
> > Jesse,
> >
> > I ran an iperf test (4 client server pairs) with and without the stats
> and
> > there was no noticeable dif
oh and one more thing I notice is that even if I do not call the function
somehow the flows get deleted. I am guessing the user space maintains its
own idle timeouts, not sure.
Thanks.
On Mon, May 26, 2014 at 5:03 PM, Madhu Challa wrote:
> Jesse,
>
> I ran an iperf test (4 client ser
GBytes 2.39 Gbits/sec
[ 3] 0.0-300.0 sec 80.1 GBytes 2.29 Gbits/sec
[ 3] 0.0-300.0 sec 82.3 GBytes 2.36 Gbits/sec
[ 3] 0.0-300.0 sec 80.3 GBytes 2.30 Gbits/sec
Thanks.
On Sun, May 25, 2014 at 10:34 AM, Jesse Gross wrote:
> On Sat, May 24, 2014 at 6:06 PM, Madhu Challa
>
Hi Jesse,
Thanks for the clarifications.
On Fri, May 23, 2014 at 9:52 AM, Jesse Gross wrote:
> On Thu, May 22, 2014 at 5:02 PM, Madhu Challa
> wrote:
> > I am looking at adding support for On-demand flow counters and figured
> one
> > of the benefits of doing this is re
I am looking at adding support for On-demand flow counters and figured one
of the benefits of doing this is reduced lock contention in
ovs_flow_stats_update.
If my understanding is correct, this involves processing flags
OFPFF13_NO_PKT_COUNTS
and OFPFF13_NO_BYT_COUNTS and communicating them so whe
25 matches
Mail list logo
