[vpp-dev] Very poor performance vm to vm via VPP vhostuser

Hello i setup 2 vm connected to VPP as per the guide : https://wiki.fd.io/view/VPP/Use_VPP_to_connect_VMs_Using_Vhost-User_Interface The performance looks very bad very low pps and large latencies udp pkt size 100B - throughput 500Mb average latency is 900us i have 2 PMDs threads (200% cpu) in

[vpp-dev] VPP build fail on ARM but pass in X86.

Hi, VPP Build is failing on ARM but the same is passing in X86. Logs :   CXX  connection.lo   CXX  dhcp_config.lo   CXX  dhcp_config_cmds.lo In file included from /home/ubuntu/vpp/build-data/../src/vpp-api/vom/cmd.hpp:21:0, from /home/ubuntu/vpp/build-data/../src/vpp

Re: [vpp-dev] VPP build fail on ARM but pass in X86.

Some piece of libboost appears to be AWOL. Copying Neale who can tell for sure. As a workaround / to make progress: in .../build-data/platforms/vpp.mk, disable the vom build as described. In build-root, issue “make distclean” and rebuild. # To disable the VPP object model build uncomment the fol

[vpp-dev] Does IPsec support jumbo frames?

While adding [0] Maximum Receive Rate tests, we have found that 9000 octet long frames never pass through a VPP setup which uses IPsec. Also, the previously existing crypto performance tests never used 9000B frames (and functional tests only use small ICMP packets), so the zero receive rate is per

Re: [vpp-dev] Very poor performance vm to vm via VPP vhostuser

in the show err output i see that l2-output l2-learn l2-input counters are continuously incremented : show err CountNode Reason 11l2-output L2 output packets 11l2-learnL2 learn

Re: [vpp-dev] Very poor performance vm to vm via VPP vhostuser

i dont think these are error counters - anyway very poor pps On Thu, Mar 22, 2018 at 2:55 PM, Sara Gittlin wrote: > in the show err output i see that l2-output l2-learn l2-input counters > are continuously incremented : > show err >CountNode Reason >

Re: [vpp-dev] Very poor performance vm to vm via VPP vhostuser

this is the output of: show vhost-user VirtualEthernet0/0/0 Virtio vhost-user interfaces Global: coalesce frames 32 time 1e-3 number of rx virtqueues in interrupt mode: 0 Interface: VirtualEthernet0/0/0 (ifindex 1) virtio_net_hdr_sz 12 features mask (0x): features (0x58208000)

Re: [vpp-dev] VPP - IPSec offloading

Hi Avi, Not exactly. VPP doesn’t use the sample app. It uses the DPDK cryptodev support, which is probably also used by the sample app. There may be a wiki page somewhere, but the doc I usually look at is in the vpp repo at src/vpp/src/plugins/dpdk/ipsec/dpdk_crypto_ipsec_doc.md -Matt > On

Re: [vpp-dev] VPP - IPSec offloading

Thank you Matt Avi > -Original Message- > From: vpp-dev@lists.fd.io [mailto:vpp-dev@lists.fd.io] On Behalf Of Matthew > Smith > Sent: Thursday, 22 March, 2018 4:22 PM > To: vpp-dev@lists.fd.io > Cc: vpp-dev@lists.fd.io > Subject: Re: [vpp-dev] VPP - IPSec offloading > > Hi Avi, > > Not e

Re: [vpp-dev] Very poor performance vm to vm via VPP vhostuser

Sara, Could you please try again with adding below config to startup.conf? I remember some drivers don’t blast traffic at full speed unless it is getting a kick for each packet sent. I am not sure if you are using one of those. vhost-user { coalesce-frame 0 } Steven From: on behalf of Sara

Re: [vpp-dev] VPP - IPSec offloading

Matt This is cryptodev and not ipsec offload (by HW) Is there offload support ? Avi > -Original Message- > From: Avi Cohen (A) > Sent: Thursday, 22 March, 2018 4:39 PM > To: 'Matthew Smith'; vpp-dev@lists.fd.io > Subject: RE: [vpp-dev] VPP - IPSec offloading > > Thank you Matt > Avi > >

Re: [vpp-dev] Very poor performance vm to vm via VPP vhostuser

Hi Sara, Can you also send the results from show hardware and show interfaces? What are you using to test your performance. John From: on behalf of Sara Gittlin Date: Thursday, March 22, 2018 at 9:27 AM To: "vpp-dev@lists.fd.io" Subject: Re: [vpp-dev] Very poor performance vm to vm via VPP

Re: [vpp-dev] Very poor performance vm to vm via VPP vhostuser

Hi Sara, To address your question on “show err”, your output is displaying running counts for the l2-output, l2-learn and l2-input node wrt how many packets each node processed. These are not error counts, unless the “Reason” column indicate some kind of error. The CLI name “show err” is mis-

Re: [vpp-dev] Very poor performance vm to vm via VPP vhostuser

Hi John and Steven setting this in the startup config didnt help vhost-user { coalesce-frame 0 } John I'm using ping -f for latency and iperf3 for pps testing. later i'll run pktgen in the vm's output : sho int Name Idx State Counter Count Vi

[vpp-dev] VPP 18.01.2 Release scheduled for Thursday March 29

Folks, Per the discussion at this week's VPP Bi-weekly Meeting, there have been enough bug fixes committed to VPP stable/1801 since the release of VPP 18.01.1 to merit another maintenance release for VPP 18.01. VPP 18.01.2 Release is scheduled for next Thursday March 29, 2018. This is a main

Re: [vpp-dev] Very poor performance vm to vm via VPP vhostuser

You should not use iperf for pps because you’re going to incur the overhead of running a guest user space app over socket + guest kernel tcp-ip stack + virtio over vhost user (times 2 since both your end points run iperf). You might get decent throughput with large write buffer size (e.g. 64k wr

Re: [vpp-dev] Does IPsec support jumbo frames?

Hi Vratko, During encapsulation/encryption, there is not currently a way to check that the buffer will not exceed the MTU of the outbound interface once padding & headers/footers are added. The outbound interface isn’t determined until after the encrypt & encapsulate is completed. If the encap

Re: [vpp-dev] Very poor performance vm to vm via VPP vhostuser

Sara, Iperf3 is not blasting traffic fast enough. You could try specifying multiple parallel streams using -P. Then, you will likely encounter vhost-user dropping packets as you are using the default virtqueue size 256. You’ll need to specify rx_queue_size=1024 when you launch the VM in the qem

Re: [vpp-dev] Does IPsec support jumbo frames?

> open a JIRA issue for it so that its tracked Opened https://jira.fd.io/browse/VPP-1207 Vratko. From: vpp-dev@lists.fd.io On Behalf Of Matthew Smith Sent: Thursday, 2018-March-22 16:45 To: vpp-dev@lists.fd.io Cc: vpp-dev@lists.fd.io Subject: Re: [vpp-dev] Does IPsec support jumbo frames? Hi

Re: [vpp-dev] make TEST=test_ip6 test failing on multiple machines, when NOT run as root.

On Tue, Mar 20, 2018 at 9:44 AM, Jon Loeliger wrote: > On Tue, Mar 20, 2018 at 6:30 AM, Ole Troan wrote: > >> >> > >> > git clean -d -f -x . >> > and rebuilding everything fixes the issue >> > >> > Let me know if you see a different issue... >> > > Ooo, I see... So, yeah, I have to hop between

Re: [vpp-dev] make TEST=test_ip6 test failing on multiple machines, when NOT run as root.

> > Ooo, I see... So, yeah, I have to hop between a local branch and the > upstream master, > so I may be being victimized by this problem. I've been doing a "make > wipe" or "wipe-release" > and forgot about the industrial 'git clean'. Drat. I'll give that a > run... > >

Re: [vpp-dev] make TEST=test_ip6 test failing on multiple machines, when NOT run as root.

On machines that I see test failures -- sometimes, frequently, they disappear when I use make test-debug rather than make test. Burt On Thu, Mar 22, 2018 at 2:08 PM, Klement Sekera wrote: > > > > Ooo, I see... So, yeah, I have to hop between a local branch and > the > > upstream master,

[vpp-dev] Is IRB feature fully functional?

Hi, I am trying to configure IRB/BVI feature and see if it works. I am doing the configuration as per the instructions given in ‘https://wiki.fd.io/view/VPP/Command-line_Interface_(CLI)_Guide#IRB.2FBVI’. Instead of configuring a route as per the link, I am configuring an IP address on BVI inter

[vpp-dev] Is IRB feature fully functional?

Hi, I am trying to configure IRB/BVI feature and see if it works. I am doing the configuration as per the instructions given in ‘https://wiki.fd.io/view/VPP/Command-line_Interface_(CLI)_Guide#IRB.2FBVI’. Instead of configuring a route as per the link, I am configuring an IP address on BVI inter

[vpp-dev] Is IRB feature fully functional?

Hi, I am trying to configure IRB/BVI feature and see if it works. I am doing the configuration as per the instructions given in ‘https://wiki.fd.io/view/VPP/Command-line_Interface_(CLI)_Guide#IRB.2FBVI’. Instead of configuring a route as per the link, I am configuring an IP address on BVI inter

Re: [vpp-dev] Is IRB feature fully functional?

In order to use a sub-interface on a loopback interface as BVI of a BD, the sub-interface should be placed in a BD and the IP address configured on the sub-interface (in your case, sub-interface with VLAN tag 50 and not the main interface). The sub-interface must be configured as an exact match

Re: [**EXTERNAL**] RE: [vpp-dev] Is IRB feature fully functional?

Thanks John for quick response. Yes. I placed sub-interface in the BD that the BVI is associated. My intention is to keep L2 functionality (vlan edits, etc) intact and add L3 to it. So this is what I am doing. vpp# loopback create-interface loop0 vpp# set interface state loop0 up vpp# set int

[vpp-dev] [FD.io Helpdesk #53548] issues with opensuse and centos verify job for stable/1801

Matus, I believe the vpp-dpdk-devel package error can be resolved by doing a rebase. Thank you, Vanessa On Tue Mar 13 11:13:13 2018, matfa...@cisco.com wrote: > Hi, > > I see issues with opensuse and centos verify jobs for stable/1801 > branch. > The error is same for both jobs: > 11:39:26

Re: [**EXTERNAL**] RE: [vpp-dev] Is IRB feature fully functional?

I read your original email as adding a sub-interface on loopback into BD as BVI. I see now you are using loop0 main interface as BVI for BD 5 while adding a sub-interface GigabitEthernet5/0/0.50 in BD 5. In this situation, it does not work because packets from sub-interface will send packets wi

Re: [**EXTERNAL**] RE: [vpp-dev] Is IRB feature fully functional?

Thanks John. That actually helped. Now I can have multiple unique ‘sub-interface:BD:BVI:VRF’s on a single physical interface. Regards, Leela sankar From: "John Lo (loj)" Date: Thursday, March 22, 2018 at 2:22 PM To: Leela Gudimetla , "vpp-dev@lists.fd.io" Subject: RE: [**EXTERNAL**] RE: [vpp-

Re: [vpp-dev] Error while running TCP_ECHO in standalone mode

The error says it all. You don’t have the session layer enabled. Do “session enable” from VPP’s cli. Florin > On Mar 22, 2018, at 8:57 AM, Shashi Kant Singh wrote: > > Hi > > I could verify TCP_ECHO as the VPP CLI app but when I run as standalone linux > app then I get this error: > > > [roo

Re: [vpp-dev] UDP ECHO Server

Your uri asks for tcp not udp transport service. Florin > On Mar 22, 2018, at 8:25 AM, Shashi Kant Singh wrote: > > Hi > > In the VPP Rel 18:01, I could run the TCP ECHO server as > vpp# test tcp server uri tcp://192.168.180.31/5678 > > But I could not find the udp server being shown under “c

Re: [vpp-dev] UDP ECHO Server

Hi Florin I want to run UDP Server. So I need to the following as I understand: vpp# test udp server uri udp://192.168.180.31/5678 Is this correct? If so, then, it does not give “udp” when I search under test cli Regards Shashi From: vpp-dev@lists.fd.io [mailto:vpp-dev@lists.fd.io] On Behal

Re: [vpp-dev] UDP ECHO Server

The API was changed some months ago to: “test echo server uri ://ip/port Let me know if it works. Thanks, Florin > On Mar 23, 2018, at 7:59 AM, Shashi Kant Singh wrote: > > Hi Florin > > I want to run UDP Server. So I need to the following as I understand: > > vpp# test udp server uri udp:/

Re: [vpp-dev] UDP ECHO Server

vpp# show version vpp v18.01.1-14~g308ed5c-dirty built by root on vbbubng6-shashi-8.rma.lan at Tue Mar 20 11:08:34 UTC 2018 vpp# test bier classify counters dpdk fib fib-walk-process frame-queue heap-validate http ip6

Re: [vpp-dev] UDP ECHO Server

Ow, I didn’t notice you were looking for it in 18.01. We didn’t have a builtin udp echo app then, that’s why the tcp one has been refactored and renamed to “test echo server/client”. Florin > On Mar 23, 2018, at 8:04 AM, Shashi Kant Singh wrote: > > vpp# show version > vpp v18.01.1-14~g308ed

Re: [vpp-dev] UDP ECHO Server

What would be the branch I can try this? Regards Shashi From: Florin Coras [mailto:fcoras.li...@gmail.com] Sent: 23 March 2018 11:52 To: Shashi Kant Singh Cc: vpp-dev@lists.fd.io Subject: Re: [vpp-dev] UDP ECHO Server Ow, I didn’t notice you were looking for it in 18.01. We didn’t have a builti

Re: [vpp-dev] UDP ECHO Server

Try master. Florin > On Mar 23, 2018, at 8:23 AM, Shashi Kant Singh wrote: > > What would be the branch I can try this? > > Regards > Shashi > > > From: Florin Coras [mailto:fcoras.li...@gmail.com > ] > Sent: 23 March 2018 11:52 > To: Shashi Kant Singh > Cc: vpp

Re: [vpp-dev] UDP ECHO Server

Thanks. Another question. I see the sock_test_server.c and sock_test_client.c for TCP alone. Is there any app for UDP on similar lines (using vpp session apis)? If not, then I can change socket to UDP and rest would work? Regards Shashi From: vpp-dev@lists.fd.io [mailto:vpp-dev@lists.fd.io] On