Re: [vpp-dev] vpp with dpdk shared library

Dear Michael, interesting work. Do you plan to upstream your PMD to DPDK mainline? Thanks, -- Damjan > On 25 May 2018, at 10:50, Michael Lilja wrote: > > Hi, > > I had to use another DPDK source base also as shared library, so I went > though a few trails before I got

Re: [vpp-dev] gdb break point of plugin not hit

Do you see in "show run" output that sample plugin node was hit with some packets? also, I suggest moving to newer version of vpp... -- Damjan > On 25 May 2018, at 21:12, Harish Patil wrote: > > Hi, > > I built VPP using make -j build TAG=vpp_debug with export >

Re: [vpp-dev] containers interconnect via memif

Ed, there are some use-cases , for example: - 2 work loads under same tenant - need to exchange info - CNF's implementing SFC Thank you -Sara On Thu, May 24, 2018 at 7:12 PM, Ed Warnicke wrote: > Sara, > > A couple question: > > What are you wanting to connect via memif in

Re: [vpp-dev] containers interconnect via memif

Thank you Damjan see setup of 2 containers one act as a master the other slave - here: https://docs.fd.io/vpp/17.10/libmemif_example_setup_doc.html but this works only with the specific image 'libmemif-sample-service' Regards -Sara On Sun, May 27, 2018 at 12:48 PM, Damjan Marion

Re: [vpp-dev] Q on build failure and memory requirement

Dear Ravi, Looks like you are hitting two different issues, but there is no enough details in your email to b able to help you. For avx2 compilation issue, please make sure that you use clean repo (i.e git clean -fdx). Also provide OS version and gcc version. On memory issue, I'm afraid that

Re: [vpp-dev] containers interconnect via memif

Sara, libmemif currently can act only as memif slave, but there is no reason for not being master, somebody just needs to extend the lib. Thanks, -- Damjan > On 27 May 2018, at 09:02, Sara Gittlin wrote: > > Ed, > there are some use-cases , for example: > - 2 work

Re: [vpp-dev] dot1ad tag

What you observed is the expected behavior. Adding such a check will add an overhead on l2-output with VTR configured and affect the packet forwarding efficiency. I still don’t understand the purpose of pushing dot1ad tag on packet at customer interface input in your VPLS test setup. It

Re: [vpp-dev] dot1ad tag

Forgot to mention, if you are willing to accept the overhead of checking the right VLAN tag is present before VTR, there is an L2 output feature called efp-filter you can enable on an interface: set interface l2 efp-filter GigabitEthernet0/8/0.200 This is typically done on a sub-interface so

[vpp-dev] Q on memif between VPP running in VM and icmp_responder in a container

Hi, I have VM (x86_64, Ubuntu 16.04) with 4GB RAM and 2 cores on which I have vpp compiled, installed and running fine. I have followed the example from https://docs.fd.io/vpp/17.10/libmemif_example_setup_doc.html Unfortunately it is not working for me. Couple of questions (1) how to debug

Re: [vpp-dev] dot1ad tag

Hi, The patch makes it work perfectly. Now the Second tag (SPtag) is removed. But I think there is another bug here. First scenario: PE1 : set interface l2 tag-rewrite GigabitEthernet5/0/0 push dot1ad 200 PE2 : set interface l2 tag-rewrite GigabitEthernet4/0/1 push dot1ad 200 On Sat, May 26,

Re: [vpp-dev] dot1ad tag

Hi, The patch makes it work perfectly. Now the Second tag (SPtag) is removed. But I think there is another bug here. *First scenario:* PE1 : set interface l2 tag-rewrite GigabitEthernet5/0/0 push dot1ad 200 PE2 : set interface l2 tag-rewrite GigabitEthernet4/0/1 push dot1ad* 200*

Re: [vpp-dev] containers interconnect via memif

Hi Sara, Please forward your original question to Jakub Grajciar: jgraj...@cisco.com He is the author of the library, who has also prepared the docker containers for the demonstration. Originally it was possible to connect two instances of libmemif directly (i.e. without VPP), but recently

[vpp-dev] VPP Vnet crash with vhost-user interface

Hello, I have a VM(16.04.4 Ubuntu x86_64) with 2 cores and 4G RAM. I have installed VPP successfully on it. Later I have created vhost-user interfaces via create vhost socket /var/run/vpp/sock1.sock server create vhost socket /var/run/vpp/sock2.sock server set interface state

Re: [vpp-dev] vpp with dpdk shared library

Hi Damjan, Well, with regards to upstreaming our PMD it is not as easy as it seems. The driver we have today is binary shipped and open-sourcing that is not just something you do over-night. We tried to upstream a PMD that would dlopen() our binary library and link into DPDK that way (this is