Pradeep
I have tried few tests by trying to turn on TSO , but according to my
observation, ovs-dpdk doesn't support TSO Offloading.
You can try this reference from the following page:
http://dpdk-test-plans.readthedocs.io/en/latest/vhost_tso_test_plan.html
Thanks
On Tue, May 22, 2018 at 1:39
Hi users,
I need to get the average time spent in packet extraction then in first
level cache , second level cache. I don't know the way to measure that.
Can anyone please help me pointing to right direction?
I think, I need to modify some code. Please guide me if I am right or wrong.
Looking
and packet header
extraction?
If yes, please tell me the way to do that? Which part I should perf for?
Thanks and Regards
On Wed, Mar 14, 2018 at 4:31 PM, Krish <attkillro...@gmail.com> wrote:
> Justin
>
> Thanks for telling me about perf tool. I think its a really good tool for
>
Hi Everyone
>From Debug information of EMC Processing , I can see the packets have IPs
as well as Ports. Packet information is below :
*|dpif_netdev(pmd8)|DBG|packet:
.
Thank you so much.
-- Forwarded message --
From: Krish <attkillro...@gmail.com>
Date: Mon, Apr 16, 2018 at 1:05 PM
Subject: Regarding Matching Flows in EMC Cache
To: ovs-discuss@openvswitch.org
Hi Everyone
>From Debug information of EMC Processing , I can see the packets
Hello everyone
I am testing ovs-vswitch caches time spent using intel vtune.
I disabled emc lookup using "emc-insert-inv-prob=0" but still I can see emc
lookup is not disabled and also there is insertion which takes place into
emc also after the packet completes fast-path processing.
I am
nk You
On Fri, Jan 11, 2019 at 2:42 AM Ben Pfaff wrote:
> On Thu, Jan 10, 2019 at 02:34:47PM +0800, Krish wrote:
> > Is there any other way to distinguish different types of packets/flows
> when
> > they enter dp_netdev_input__ and processed further(EMC,fastpath etc)?
> >
Is there any other way to distinguish different types of packets/flows when
they enter dp_netdev_input__ and processed further(EMC,fastpath etc)?
As of now, I have to *flow_extract* the packets using *struct flow flow*:
if (flow.nw_proto == IPPROTO_TCP && l4_size >= TCP_HEADER_LEN) I want to
You don't need to give ip address to ovs bridge.
As you VM's are already connected through ovs-bridge, so you can simple use
iperf server on VM1 and client on another (VM2). In this case, ovs is only
working as a bridge to pass traffic from one Vm to another(as they are
already connected).
Thanks
gt; 20~30 ms (32 packets)
Can anyone please throw some light on this?
Waiting for a response. Thank You
Krish
___
discuss mailing list
disc...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-discuss
Sorry , I forgot to mention I am using dpdk - 17.05 and OVS - 2.8.1
Thanks
Krish
On Mon, Mar 25, 2019 at 2:07 PM Krish wrote:
> Hi
>
> I am wondering why does batch execution takes more time than EMC and
> fast_path processing even if there are 32 packets grouped together from a
w_proto(ip protocol) is printing out to be 0.
I am using this code just before emc_processing and inside
dp_netdev_input__.
Can you please help me here?
Thanks
Krish
On Mon, Feb 18, 2019 at 8:11 PM Flavio Leitner wrote:
> On Mon, Feb 18, 2019 at 11:05:53AM +0800, Krish wrote:
> > Thanks for
c));
VLOG_DBG( "tcp_dst ---> %d", ntohs(th->tcp_dst));
}
}
Can you please confirm, if this is the right approach I have used and not
seeing the old values like before?
Thank you so much,
Krish
On Tue, Feb 19, 2019 at 11:25 AM Krish wrote:
> Flavio,
>
> Thank you
Hey Flavio
I am so sorry for the inconvenience caused.
Next time I will take care of it. You have been a great help for me.
I thank you from the bottom of my heart. I was stuck on this part since
last few days.
Now, it has been resolved.
Thanks
Krish
On Wed, Feb 20, 2019 at 8:25 PM Flavio
Hey
This might help you.
Take a look please
https://mail.openvswitch.org/pipermail/ovs-discuss/2016-March/040196.html
Thanks
On Wed, Feb 6, 2019, 18:45 장예훈 wrote:
> Hi.
>
> I’m analyzing the source code of ovs for first time.
>
>
>
> I’d like to find a source code that handles packets
and
print it in vlogs?
Statement 2 is working fine if vlogs are disabled, but I want the port
value to be stored in a variable and then print.
Thanks
Krish
___
discuss mailing list
disc...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-discuss
ong?
Thanks a lot for your patience and response.
Regards
Krish
On Sat, Feb 16, 2019 at 10:09 AM Ben Pfaff wrote:
> You are almost certainly dereferencing a null pointer.
>
> On Sat, Feb 16, 2019 at 09:02:17AM +0800, Krish wrote:
> > That code I showed before. It is working fine and
I am sorry, but I am trying not to use extracted flow. I want to use this
information before it goes to EMC.
On Sat, Feb 16, 2019, 03:44 Ben Pfaff wrote:
> On Fri, Feb 15, 2019 at 03:48:22PM +0800, Krish wrote:
> > Hi
> >
> > I want to store tcp_src, dst port in a varia
Hi
Yes, you can modify the ovs code as per your requirements but before that
you need to understand and gain more knowledge about its mechanism and
study the call graphs, how the code flows. Once you understand this, it
will be easier for you.
For debugging packets, you can make use of VLOG
Why does batch execution takes more time than EMC and fast_path processing
even if there are 32 packets grouped together from a same flow.
I am using iperf as a packet generation testing tool.
EMC processing -> 2~3 ms (32 packets)
Fast path -> .3 ms
packet_batch_per_flow_execute -> 20~30 ms (32
Hi
Why does batch execution takes more time than EMC and fast_path processing
even if there are 32 packets grouped together from a same flow.
I am using iperf as a packet generation testing tool.
EMC processing -> 2~3 ms (32 packets)
Fast path -> .3 ms
packet_batch_per_flow_execute -> 20~30 ms
In my OVS-DPDK setup with VxLAN tunnel between two devices, I see the traffic stops
when I flush the ARP cache using 'ovs-appctl tnl/arp/flush'. On the packet capture, I
see the device on which I flushed the arp cache initiates the ARP request and also
receives the ARP response but it doesn't
22 matches
Mail list logo