Re: [vpp-dev] Insert Custom Node Into ip4-unicast Feature Arc

Hi Neale, On Sat, Oct 1, 2022 at 1:10 AM Neale Ranns wrote: > > > Hi Padraig, > > > > *From: *vpp-dev@lists.fd.io on behalf of Pádraig > Connolly via lists.fd.io > *Date: *Saturday, 1 October 2022 at 09:52 > *To: *vpp-dev@lists.fd.io > *Subject: *[vpp-dev] Insert Custom Node Into ip4-unicast

Re: [vpp-dev] Insert Custom Node Into ip4-unicast Feature Arc

Hi Padraig, From: vpp-dev@lists.fd.io on behalf of Pádraig Connolly via lists.fd.io Date: Saturday, 1 October 2022 at 09:52 To: vpp-dev@lists.fd.io Subject: [vpp-dev] Insert Custom Node Into ip4-unicast Feature Arc Hi, For purely my own learning, I am trying to insert my custom node into

[vpp-dev] Insert Custom Node Into ip4-unicast Feature Arc

Hi, For purely my own learning, I am trying to insert my custom node into the ip4-unicast feature arc. Based on what I know so far, I would expect the following FEATURE_INIT struct to achieve what I want : VNET_FEATURE_INIT (padraig, static) = { .arc_name = "ip4-unicast", .node_name =

[vpp-dev] "vpp_agent vpp_papi.vpp_papi.VPPIOError: [Errno 2] Connect failed" seen with networking-vpp bringup on openstack #vpp-dev

I have Ubuntu 20.04 with AIO openstack installed from devstack.  Openstack deployment came up fine but VPP-agent is throwing below error. Any help would be appreciated I was following the steps mentioned here https://opendev.org/x/networking-vpp/src/branch/master stack@gnaosp001:~/devstack$

Re: [vpp-dev] nat44-ed distinct pool for out if

You are very welcome. Best regards, Filip Varga st 28. 9. 2022 o 16:32 Amir Hossein <1377amirhossein1...@gmail.com> napísal(a): > Oh you're right. > thanks to your guidance i was able to reach correct configuration. > thanks a lot! > > > -=-=-=-=-=-=-=-=-=-=-=- Links: You receive all

[vpp-dev] Documentation to install networking VPP plugin on openstack

HI ALL, Can someone help and point me to the documentation to install networking VPP plugin on openstack. thanks, Suresh -=-=-=-=-=-=-=-=-=-=-=- Links: You receive all messages sent to this group. View/Reply Online (#21947): https://lists.fd.io/g/vpp-dev/message/21947 Mute This Topic:

Re: [vpp-dev] Stats (rx/tx bytes) per MAC address

Great. Thanks for taking the time to reply. Andre Courchesne On Wed, Sep 28, 2022 at 7:04 PM Neale Ranns wrote: > Hi Andre, > > > > The adjacency counters are for TX. There’s no way currently to get RX > stats per MAC. > > > > /neale > > > > *From: *vpp-dev@lists.fd.io on behalf of Andre >

Re: [vpp-dev] Stats (rx/tx bytes) per MAC address

Hi Andre, The adjacency counters are for TX. There’s no way currently to get RX stats per MAC. /neale From: vpp-dev@lists.fd.io on behalf of Andre Courchesne via lists.fd.io Date: Thursday, 29 September 2022 at 04:00 To: vpp-dev@lists.fd.io Subject: Re: [vpp-dev] Stats (rx/tx bytes) per

Re: [vpp-dev] Stats (rx/tx bytes) per MAC address

Thanks Neale. This seems to provide only total packet/bytes and does not split between RX and TX. Know any way to get the discrete RX and TX counters ? Andre Courchesne - Consultant http://net-forces.blogspot.ca https://www.co2.click Twitter: @IndianaTux LinkedIn:

Re: [vpp-dev] nat44-ed distinct pool for out if

Oh you're right. thanks to your guidance i was able to reach correct configuration. thanks a lot! -=-=-=-=-=-=-=-=-=-=-=- Links: You receive all messages sent to this group. View/Reply Online (#21943): https://lists.fd.io/g/vpp-dev/message/21943 Mute This Topic:

[vpp-dev] MAC address only based ACL

Hi, I have read many documentation but did not find my answers, so here is the question. Is there a way in VPP to redirect all TCP traffic to a host Linux process unless the traffic comes from a MAC address that is part of some sort of ACL. We used to do this without VPP using iptables and

Re: [vpp-dev] There is bug in patch

I have some idea that is not good 1, spi should be one key in inbound policy serach typedef union { struct { ip4_address_t ip4_src_addr; ip4_address_t ip4_dest_addr; U32 spi; ipsec_spd_policy_type_t policy_type; }; // 16 bytes total

Re: [vpp-dev] There is bug in patch

Looks like it would be a good time to unify slow-path/fastpath and inbound/outbound policy matching code  Best ben > -Original Message- > From: vpp-dev@lists.fd.io On Behalf Of Zachary Leaf > Sent: Wednesday, September 28, 2022 10:10 > To: vpp-dev@lists.fd.io > Cc: Zhang, Roy Fan ;

Re: [vpp-dev] There is bug in patch

Copying in Piotr, Fan and some other folks for awareness/any input. Best, Zach On 28/09/2022 09:07, Zachary Leaf via lists.fd.io wrote: > Hi Guangming, > > Thanks for the report. I think you may have uncovered some of the bugs > in the general inbound matching logic. > > I think there are 2

Re: [vpp-dev] There is bug in patch

Hi Guangming, Thanks for the report. I think you may have uncovered some of the bugs in the general inbound matching logic. I think there are 2 problems with the logic for the standard linear search (not flow cache, but impacts flow cache): 1. Only matches source/dest ip, and doesn't match

Re: [vpp-dev] Stats (rx/tx bytes) per MAC address

Hi Andre, For L3 forwarding you can use the per-adjacency counters. This are not enabled by default, because, like all stats, they come with a performance cost. DBGvpp# adjacency counters ? adjacency counters adjacency counters [enable|disable] you’ll then find then

[vpp-dev] Stats (rx/tx bytes) per MAC address

Hi, Is there a way to retrieve the number of bytes per MAC address transferred through VPP ? Thanks. Andre -=-=-=-=-=-=-=-=-=-=-=- Links: You receive all messages sent to this group. View/Reply Online (#21936): https://lists.fd.io/g/vpp-dev/message/21936 Mute This Topic:

[vpp-dev] 2022-09-27 FD.io VPP (master branch) Gerrit Change Report

New report generated today: https://gist.github.com/dwallacelf/4147e4d26ca154698ea3c881426160c3 Please review the report and update/review/merge changes as appropriate. Thanks, -daw- On 9/14/22 3:55 PM, Dave Wallace via lists.fd.io wrote: New report generated today:

Re: [vpp-dev] input_policy_epoch_by_sw_if_index what is significance of this epoch ? #acl_plugin

it would depend on what are you looking to achieve ? --a > On 27 Sep 2022, at 07:00, gopi.puligun...@gmail.com wrote: > > code is quite tough to understand as not sure of use cases and less > documentation , and is taking time as a new enthusiast, > any suggestion what's better way to

Re: [vpp-dev] input_policy_epoch_by_sw_if_index what is significance of this epoch ? #acl_plugin

code is quite tough to understand as not sure of use cases and less documentation , and is taking time as a new enthusiast, any suggestion what's better way to approach , any specific resources or documentation you recommend Andrew ? -=-=-=-=-=-=-=-=-=-=-=- Links: You receive all messages sent

[vpp-dev] 'make test' failures on ubuntu2204 & debian11

Folks, Due to the upgrade to ubuntu-22.04 and re-enabling multi-worker tests on debian-11 CI jobs, there have been a number of regressions in 'make test' (some intermittent failures, other hard failures). I have disabled the offending tests [0] and rebased all of the outstanding gerrit

[vpp-dev] There is bug in patch

Hi, Zach I think there is a bug in patch : perf improvement of ipsec4_input_node using flow cache (https://gerrit.fd.io/r/c/vpp/+/32903). Based on current code , multiple spd rules were created that include some bypass or discard rule . When one ESP packet was received , we expect it

[vpp-dev] #vpp-dev: VPP support on RedHat Openstack platform

Is VPP supported on RedHat Openstack platform ? -=-=-=-=-=-=-=-=-=-=-=- Links: You receive all messages sent to this group. View/Reply Online (#21930): https://lists.fd.io/g/vpp-dev/message/21930 Mute This Topic: https://lists.fd.io/mt/93936232/21656 Group Owner: vpp-dev+ow...@lists.fd.io

Re: [vpp-dev] UDP multithreading error

Hi, That’s an inefficiency in ldp when vcl does not use eventfds for notifications. Could you change your vcl.conf and add use-mq-eventfd? Regards, Florin > On Sep 26, 2022, at 5:35 AM, nengjie wang wrote: > > Thank you very much for your reply. > > However, after the socket is set to the

Re: [vpp-dev] Mellanox Interface not coming up in VPP 22.02

Hoi, /sub (as I have had the same issue when testing Cx6 cards on Epyc boards a while ago. I never got them to work with DPDK, although I spent many evenings trying. Chinmaya - obvious question: I assume that you have tried and ruled out using the more widely used RDMA driver for the Mellanox

Re: [vpp-dev] UDP multithreading error

Thank you very much for your reply. However, after the socket is set to the non blocking mode, the CPU utilization rate is always 100%. The resource consumption is very high when multithreading. I can only reduce the CPU utilization rate by using methods such as select. Why does this mode fail

[vpp-dev] create host-interface failed

Hi, Why do I always return invalid argument when I execute the command create host interface name vpp? Does af packet v3 have any special dependencies? My vpp version is 22.06, the system version is centos7, the GLIBC version is 2.17, and the kernel version is 3.10.0-1127. Error message:

[vpp-dev] Mellanox Interface not coming up in VPP 22.02

Hi, We have compiled VPP v22.02 with mellanox (MLNX_OFED_LINUX-5.5-1.0.3.2-rhel8.4-x86_64) on Centos 8.4 VM. We are able to configure interface (having mellanox driver) inside VPP but when we try bringing up the interface using "set interface state..", it shows the error "Interface start

Re: [vpp-dev] input_policy_epoch_by_sw_if_index what is significance of this epoch ? #acl_plugin

Lazily reclassifying the sessions which were created under a different policy and thus potentially are not allowed anymore. --a > On 26 Sep 2022, at 12:24, gopi.puligun...@gmail.com wrote: > > could you please provide some info on input_policy_epoch_by_sw_if_index > purpose ? > >

[vpp-dev] input_policy_epoch_by_sw_if_index what is significance of this epoch ? #acl_plugin

could you please provide some info on input_policy_epoch_by_sw_if_index purpose ? -=-=-=-=-=-=-=-=-=-=-=- Links: You receive all messages sent to this group. View/Reply Online (#21923): https://lists.fd.io/g/vpp-dev/message/21923 Mute This Topic: https://lists.fd.io/mt/93923737/21656 Mute

Re: [vpp-dev] Unable to access ipsec_sa_t object from my own plugin ( non esp plugins)!

Hi All, Issue is solved after linking the below files to my custom plugin. +include_directories(/src/vnet) +include_directories(/src/vnet/ipip) +include_directories(/src/vnet/ipsec) Also I did extern ipsec_sa_t * in my plugin. Thanks, Sagar On Thu, Sep 22, 2022 at 10:29 PM sagar g

[vpp-dev] VPP PPPOE and PPP stack is available or not? #hoststack

Hi , I am looking for VPP release with PPPoE and PPP stack that is supported.. It is PPPoE server. WR, Sunder -=-=-=-=-=-=-=-=-=-=-=- Links: You receive all messages sent to this group. View/Reply Online (#21921): https://lists.fd.io/g/vpp-dev/message/21921 Mute This Topic:

Re: [vpp-dev] testing vpp question

Hello, I was using an unstable base version, but it's fine now. Thanks for the advise, Laszlo -=-=-=-=-=-=-=-=-=-=-=- Links: You receive all messages sent to this group. View/Reply Online (#21920): https://lists.fd.io/g/vpp-dev/message/21920 Mute This Topic: https://lists.fd.io/mt/93872230/21656

Re: [vpp-dev] nat44-ed distinct pool for out if

"Using vrf routes and tables affect selected tx_fib_index but address allocation for snat(nat_ed_alloc_addr_and_port function) is based on rx_fib_index." - Pretty sure that's not true at all I would know as a maintainer of nat :). If tx_fib_index is supplied then resolution of the pool address is

Re: [vpp-dev] nat44-ed distinct pool for out if

thanks for your detail answer. but again my goal is to have separate nat ip pool per output interface. in my case i have one in interface and two out if, i want out if1 uses nat ip address A  and out if2 uses nat ip address B. using vrf routes and tables affect selected tx_fib_index but address

Re: [vpp-dev] testing vpp question

Hi Laszlo, Are you trying to run specifically the same tests as ones that are in jenkins ? You can simply install all requirements make install-deps make install-ext-deps make test or make test-debug I would also suggest running it in 22.04 ubuntu docker image. Best regards, Filip Varga pi

[vpp-dev] testing vpp question

Hi, Could someone please advise me how to run vpp tests? I followed the instructions described in developer documentation https://s3-docs.fd.io/vpp/22.10/developer/build-run-debug/index.html I also checked the following Jenkins job configuration:

[vpp-dev] have a common session managment db which all nodes can use (either core plugins or product specific plugins) can leverage by avoiding redundancy of 5-tuple #acl

Hi All, one observation in relation to session management in acl-plugin and other plugins is, each plugin/node are maintaining different session database which has mostly redundant entries and is memory intensive. as for session management 5tuple is common .. it would be good if vpp provide a

[vpp-dev] Unable to access ipsec_sa_t object from my own plugin ( non esp plugins)!

Hi All, I am trying to do the following tasks while deleting the sa to handle inflight packets. 1. sa delete event processed from main thread 2. I am trying to delay sa delete operation and queue this event to maintenance list 3. There is a timer run in each worker thread which iterates through

Re: [vpp-dev] NAT44 for TCP/UDP

Hi Ashwini, I am sorry I wasn't able to respond earlier. Basically you have multiple issues here. You are not configuring NAT outside address as suggested by yangbin. Though you can ignore the default route comment you don't need to worry about this to make the example work. Also output-feature

Re: [vpp-dev] nat44-ed distinct pool for out if

Hi Amir, If I understand your use case correctly you have 2 inside and 2 outside interfaces. 2 outside belong to two different ISPs. You want communication entering inside interface one to go through let's say outside interface one and vice versa for the other two. inside 1 <-> outside 1 inside

Re: [vpp-dev] nat44-ed distinct pool for out if

Hi Filip, thanks for suggestion, so i can have a vrf(rx) use nat ip pool from another tenant vrf? I'm a little confused about using nat vrf routes can you give me an example? -=-=-=-=-=-=-=-=-=-=-=- Links: You receive all messages sent to this group. View/Reply Online (#21911):

Re: [vpp-dev] nat44-ed distinct pool for out if

Hi Ole , thanks for hint, but i can't see how using output-feature helps me to have different pool for out interface?(a->fib_index == rx_fib_index is in tx_sw_if_index != ~0 too) can you elaborate more pls? Thanks. -=-=-=-=-=-=-=-=-=-=-=- Links: You receive all messages sent to this group.

Re: [vpp-dev] VPP routing rule installation limit #vpp_papi #vapi

Thanks for all your help. The problem is the memory limit in the startup configuration file. Specifically, the size of the `stat-segment,` like Pim mentioned, could be an affecting factor. Besides that, the memory size of the main heap should also be increased. The resulting configuration

[vpp-dev] VPP 22.10 RC1 milestone is complete!

Hi all, VPP 22.10 RC1 milestone is complete! The artifacts for 22.10 RC1 are available at packagecloud.io/fdio/2210 Now the master branch is open for all commits, and the new stable/2210 - for low risk changes and fixes in preparation for the 22.10 release. All changes destined to stable/2210

Re: [vpp-dev] NAT44 for TCP/UDP

There are 2 types of NAT implementation. when you use "in" and "out" at same time. it's PRE-ROUTING NAT when you use "out" with "output-feature" keyword . it's POST-ROUTING NAT. no matter what type of NAT  you choose to use, there are 2 import thing you need to do, first,enable nat second,make

Re: [vpp-dev] NAT44 for TCP/UDP

I would advise trying command below: nat44 plugin enable sessions 63000 nat44 forwarding enable set nat timeout udp 3000 tcp-established 7440 tcp-transitory 2400 icmp 600 set interface nat44 out  {wan1} output-feature nat44 add static mapping tcp local 192.168.1.20 external 4.4.4.4 nat44 add

Re: [vpp-dev] nat44-ed distinct pool for out if

Hi, I would suggest checking vrf table implementation in NAT44-ED that enables restriction of passing packets between different VRFs for more information check commit 691c630b79ec2230b67944c8a9f77f95b49d95ad Meanwhile I will look if there is something else to use. If I find something I will

Re: [vpp-dev] nat44-ed distinct pool for out if

Hi Amir, Have you tried configuring the NAT with “output-feature” enabled? Best regards, Ole > On 21 Sep 2022, at 06:43, Amir Hossein <1377amirhossein1...@gmail.com> wrote: > >  > Hi, > I'm working with nat44-ed plugin and I found that we can set a distinct nat > ip pool for each in if

[vpp-dev] nat44-ed distinct pool for out if

Hi, I'm working with nat44-ed plugin and I found that we can set a distinct nat ip pool for each in if (with tenant-vrf) and we can have multiple out if. But is there a way to have a distinct nat ip pool for each out if? Imagine a scenario in which a host connects to two isps (with two if each

[vpp-dev] Reminder: VPP 22.10 RC1 milestone tomorrow 21 September 12:00 UTC

Hi all, Just a small reminder about our upcoming 22.10 RC1 milestone tomorrow (Wednesday 21th September). I will be cutting the stable/2210 branch at 12:00 UTC, similar to previous releases. Please hold on merging risky changes to master branch until that step is complete. Thanks a lot!

Re: [vpp-dev] IPSec with tables is not working

I suspect the issue is that you let the IKEv2 plugin create the ipip0 interface automatically for you: by default it will use table 0 to reach its peer. When you set Radio-0 in table 1, ipip0 can no longer reach its peer by looking up the next hop in table 0. Instead, try to create the ipip

Re: [vpp-dev] ISIS with VPP+FRR

VPP accepts the frame based on the MAC, but there is no handler for the ISIS MAC type, so it is dropped. You’ll need to add support for this. /neale From: vpp-dev@lists.fd.io on behalf of Mohamed Naiem via lists.fd.io Date: Thursday, 15 September 2022 at 21:08 To: vpp-dev@lists.fd.io

[vpp-dev] Input on VPP 'make test' use cases

Folks, In preparation for refactoring 'make test', I'd like to get feedback on the which features in test/Makefile are being used and the context for them (e.g. FD.io CI, downstream CI, local testing, test development). I've created a wiki page [0] to capture the feedback.  Please edit the

[vpp-dev] IPSec with tables is not working

Hi all, We want to run IPSec over vpp and used the IPsec plugin. It did work fine as long as we don’t use routing tables in VPP. Are routing tables with IPSec generally not supported in VPP or is there an error in our configuration? Any help is appreciated! VPP Version: vpp

Re: [vpp-dev] request-response between vlib processes

> why we ended up needing such a complicated procedure to apply what looked > like simple updates. All I can tell is that avf_process was preferring to suspend itself (rather than busy wait) already in the first AVF commit [13], and later the wait times got even longer [14]. > I’m pretty sure

[vpp-dev] Not able to configure iif interface as sub interface in End.AD.Flow

Hi, We are trying to test a use case in Srv6 proxy using VPP v22.02, where we want to configure a End.AD.Flow sid that has iff interface as vlan sub interface. But when we run the below command, we see the error:- vpp# sr localsid address 2001:f1:f1:11::111 behavior end.ad.flow nh 9.9.10.102

Re: [vpp-dev] Reminder: 7 days to RC1 milestone of VPP 22.10

Hi Ting, i am sorry, indeed you are right! I compared the latest change# with the previous one, not with the #16 which was there before the meeting. I was wrong with my assessment. So the final +2 word is for Dave/Damjan since the comments were theirs. From release management PoV I am happy

Re: [vpp-dev] mellanox mlx5 + rdma + lcpng + bond - performance (tuning ? or just FIB/RIB processing limit)

> Currently just playing with some settings and checking what I can show for > debug when I will test it again - because now it is a little hard to debug > when there is no info how much load traffic is doing to cpu's (for example > cpu cycles etc... something like perf top) It is the purpose of

Re: [vpp-dev] Reminder: 7 days to RC1 milestone of VPP 22.10

Hi, Andrew The latest version is the one modified according to the comments. I deleted most of the samples, left a few typical ones, and write the guidance how to create the json profile in README and Spec Thanks -=-=-=-=-=-=-=-=-=-=-=- Links: You receive all messages sent to this group.

Re: [vpp-dev] mellanox mlx5 + rdma + lcpng + bond - performance (tuning ? or just FIB/RIB processing limit)

Hi Yes - probabbly this tap interface is just only logical local only for ip_input/ip_output (local) traffic I will share show run after I will have another test window in my lab :) Currently just playing with some settings and checking what I can show for debug when I will test it again -

Re: [vpp-dev] VPP routing rule installation limit #vpp_papi #vapi

Hoi, It could be that it's the stats-seg that is running out. I found that with >1M routes in FIB, we need a bit more memory. For our routers, startup.conf has: statseg { size 1G page-size default-hugepage per-node-counters off } And with this: pim@nlams0:~$ vppctl show memory

Re: [vpp-dev] VPP routing rule installation limit #vpp_papi #vapi

Hi, I think you might find the thread in https://www.mail-archive.com/vpp-dev@lists.fd.io/msg16302.html useful. --a > On 15 Sep 2022, at 17:39, yxm...@case.edu wrote: > Hello, > > I am trying to evaluate the forwarding efficiency of VPP with large FIB/RIB, > like millions of entries. > >

[vpp-dev] VPP routing rule installation limit #vpp_papi #vapi

Hello, I am trying to evaluate the forwarding efficiency of VPP with large FIB/RIB, like millions of entries. However, the VPP crashes when inserting routing entries of more than 310k. The exact number varies across different routing rulesets but should be around 320k. (One test is 319,739,

Re: [vpp-dev] Reminder: 7 days to RC1 milestone of VPP 22.10

Hi Ting, I would classify this change as an "extremely low risk" one since everything is in ~vpp/extras - so I am happy to have the deadline for it be RC2 milestone. However, having just glanced at the patch, I see you uploaded the new revision yesterday and it changes the commit message. When do

Re: [vpp-dev] ISIS with VPP+FRR

Looks for me like problem is VPP doesn't know about AllIntermediateSystems MAC so it drops received packet, I tired to switch network-type to broadcast but still doesn't work considering AllL2ISs MAC as well. Not sure if thats right but I'm trying to let VPP know about ISIS MACs

Re: [vpp-dev] ISIS with VPP+FRR

>From trace, I see it says 'unknow osi protocol' 16:43:14:359482: llc-input LLC osi_layer5 -> osi_layer5 16:43:14:359482: osi-input OSI isis 16:43:14:359482: error-drop rx:BondEthernet0.2000 16:43:14:359482: drop *osi-input: unknown osi protocol* -=-=-=-=-=-=-=-=-=-=-=- Links: You receive all

Re: [vpp-dev] VPP crashing if we configure srv6 policy with five sids in the sidlist

I make a clarification , only with SRv6 hyperf3 does not work. with normal traffic it works correctly -=-=-=-=-=-=-=-=-=-=-=- Links: You receive all messages sent to this group. View/Reply Online (#21885): https://lists.fd.io/g/vpp-dev/message/21885 Mute This Topic:

Re: [vpp-dev] VPP crashing if we configure srv6 policy with five sids in the sidlist

Hi Pablo, I compiled VPP following the guidelines in the post you provided. Using the official github https://github.com/FDio/vpp. But the moment I start VPP I have these warnings: vpp[5924]: perfmon: skipping source 'intel-uncore' - intel_uncore_init: no uncore units found vpp[5924]:

Re: [vpp-dev] mellanox mlx5 + rdma + lcpng + bond - performance (tuning ? or just FIB/RIB processing limit)

Well, it depends how your traffic is flowing. From the output you shared, the rdma interface does have 8 rxq in poll-mode, 1 per worker using 8 workers (side note: that means you're not going to use more than 8 cores to process packets from that interface, even if you dedicated 24 cores to

Re: [vpp-dev] Reminder: 7 days to RC1 milestone of VPP 22.10

Hi, Andrew My patch https://gerrit.fd.io/r/c/vpp/+/36149 (packetforge) is waiting to be merged, thanks. Best Regards, Xu Ting -=-=-=-=-=-=-=-=-=-=-=- Links: You receive all messages sent to this group. View/Reply Online (#21882): https://lists.fd.io/g/vpp-dev/message/21882 Mute This Topic:

Re: [vpp-dev] mellanox mlx5 + rdma + lcpng + bond - performance (tuning ? or just FIB/RIB processing limit)

Hi So hmm... maybee there is some problem with rx/tx queue creation because after this commands: create interface rdma host-if enp59s0f0 name enp59s0f0-rdma rx-queue-size 4096 tx-queue-size 512 num-rx-queues 8 no-multi-seg lcp create enp59s0f0-rdma host-if e0-rdma When i do show for tap

[vpp-dev] 2022-09-14 FD.io VPP (master branch) Gerrit Change Report

New report generated today: https://gist.github.com/dwallacelf/c60164b8affb65fdef690d7500d8de1a Please review the report and update/review/merge changes as appropriate. Thanks, -daw- On 9/1/22 4:04 PM, Dave Wallace via lists.fd.io wrote: Folks, I have updated a gerrit change report

[vpp-dev] ISIS with VPP+FRR

Hi, I'm trying to configure ISIS between Cisco router and host that runs VPP and FRR for control plane. All looks good for any other routing protocol like BGP and OSPF, also LDP but not ISIS. My guessing because ISIS uses different L2 encapsulation ? Couldn't find any config knob that might

Re: [vpp-dev] FDIO Maintenance - 2022-09-14 16:30 UTC

Maintenance is complete. Thank you for your patience. Vanessa On 9/14/22 11:13 AM, Vanessa Valderrama wrote: *What:* * Jenkins production and sandbox o OS and security updates o Jenkins upgrade o Plugin upgrades *When: *2022-09-14 16:30 UTC *Why:* Jenkins upgrade due

[vpp-dev] FDIO Maintenance - 2022-09-14 16:30 UTC

*What:* * Jenkins production and sandbox o OS and security updates o Jenkins upgrade o Plugin upgrades *When: *2022-09-14 16:30 UTC *Why:* Jenkins upgrade due to multiple CVEs *Impact:* The following services will be unavailable during the maintenance window: * Jenkins

[vpp-dev] Reminder: 7 days to RC1 milestone of VPP 22.10

Hi all, We have a week left before the RC1 milestone, during which I will pull stable/2210 branch in preparation for the release. Please let me know if you have unmerged changes that you need to get in before RC1 that i can help with. Thanks ! --a /* your friendly 22.10 release manager */

Re: [vpp-dev] request-response between vlib processes

Hi Vratko, > On Sep 13, 2022, at 5:03 AM, Vratko Polak -X (vrpolak - PANTHEON TECHNOLOGIES > at Cisco) via lists.fd.io wrote: > > In general, most of “communication” between VPP components > is done by directly calling C functions, > so it makes sense avf_flag_change is being called within

Re: [vpp-dev] benchmark for a patch

> a specific patch which is not merged? It depends whether the unmerged patch is for VPP or CSIT (or you have two). 1. If VPP code is merged, various csit-verify-perf-* jobs can be triggered using csit-{node}-{arch}-perftest trigger word in CSIT Gerrit comment. Example: [0]. The VPP version used

Re: [vpp-dev] request-response between vlib processes

In general, most of “communication” between VPP components is done by directly calling C functions, so it makes sense avf_flag_change is being called within vl_api_clnt_process process. It is avf_process_request (called directly by avf_flag_change) that decides to hand-off the request to

Re: [vpp-dev] request-response between vlib processes

Hi Vratko, Do we really need to block the binary api waiting for a reply from another vpp process just to set a mac address? If setting up the mac (or similar) cannot be done synchronously, probably api handlers should hand over all those requests to another vpp process,

[vpp-dev] request-response between vlib processes

[resending to the correct vpp-dev e-mail address] Short version: Vratko would appreciate something like vlib_current_process_wait_for_one_time_event_or_clock. Medium version: One instance of request-response interaction between vlib processes had a bug [11]. Vratko contributed a fix [9] for

Re: [vpp-dev] Is BGP or OSFP support in VPP

Hi Ole, Thank you for the clarification. On Mon, 12 Sep 2022, 16:20 Ole Troan, wrote: > Vijay, > > VPP is a data plane. > There are various ways of integrating with routing protocols. The simplest > is likely to use the linux-cp plugin and use linux as the control plane. > > Best regards,

Re: [vpp-dev] Is BGP or OSFP support in VPP

Vijay, VPP is a data plane. There are various ways of integrating with routing protocols. The simplest is likely to use the linux-cp plugin and use linux as the control plane. Best regards, Ole > On 12 Sep 2022, at 08:38, Vijay Kumar wrote: > >  > Hi, > > Based on exploring the

Re: [vpp-dev] mellanox mlx5 + rdma + lcpng + bond - performance (tuning ? or just FIB/RIB processing limit)

This is what we measure on CSIT with Intel Cascade Lake for 1 CPU core:

[vpp-dev] Is BGP or OSFP support in VPP

Hi, Based on exploring the files/folder in the source code that I downloaded, it is very much clear that there is no support for dynamic routing in VPP right now. I believe right now, the only option to route traffic from VPP is by adding the static routes. Is there any plan to support BGP or

[vpp-dev] mellanox mlx5 + rdma + lcpng + bond - performance (tuning ? or just FIB/RIB processing limit)

Hi First I want to thanks to all ppl that created RDMA native support in vpp and also ppl from LCPNG / Linux-CP - it is working and looks stable :) But Was testing some scenarions with rdma+vpp+lcpng+frr BGP with 200k routes - with mellanox mlx5 2x100G nics - where 24 cores are used for

Re: [vpp-dev] vpp vpp v22.10-rc0~177-g563d34ba9 lcp create unknown input `' problem

Hoi Pawel, On Fri, Sep 9, 2022 at 2:12 AM Paweł Staszewski wrote: > Currently i stuct with other thing - can't configure more than one sub > interface as vlan all because after configuring one vlan second one always > has error: > > vpp# create sub BondEthernet0 4070 > BondEthernet0.4070 > vpp#

Re: [vpp-dev] vpp vpp v22.10-rc0~177-g563d34ba9 lcp create unknown input `' problem

ok looks like I was typing it wrong... host-if is not host-if it is some if :) Currently i stuct with other thing - can't configure more than one sub interface as vlan all because after configuring one vlan second one always has error: vpp# create sub BondEthernet0 4070 BondEthernet0.4070

Re: [vpp-dev] vpp vpp v22.10-rc0~177-g563d34ba9 lcp create unknown input `' problem

Hi So here is output of show int and int creation of rdma mellanox nic ip a 1: lo: mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000     link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00     inet 127.0.0.1/8 scope host lo    valid_lft forever preferred_lft forever    

Re: [vpp-dev] lpm table api?

I found lpm api and see the map plugin uses the api. I'm all set. https://docs.fd.io/vpp/19.04.3/d5/db5/lpm_8c_source.html Hemant From: vpp-dev@lists.fd.io On Behalf Of hemant via lists.fd.io Sent: Wednesday, September 07, 2022 9:57 AM To: vpp-dev@lists.fd.io Subject: [vpp-dev] lpm

Re: [vpp-dev] vpp vpp v22.10-rc0~177-g563d34ba9 lcp create unknown input `' problem

Hi, The help text can be displayed with '?' rather than 'help': vpp# lcp create ? lcp create lcp create | host-if netns [tun] When you ran 'lcp create 1', you got a different error message than you did with the rest of your commands. That suggests that 1 is a

[vpp-dev] vpp vpp v22.10-rc0~177-g563d34ba9 lcp create unknown input `' problem

Hi Was trying linux-cp vpp and the problem is that i cant create any interface with lcp because any command is like this: vpp# lcp create vpp# lcp create help lcp create: unknown input `' vpp# lcp create default lcp create: unknown input `' vpp# lcp create vpp# lcp create1 lcp: unknown input

Re: [vpp-dev] UDP multithreading error

Hi, Just tested on master and this seems to work fine after a few app fixes. Having said that, I would recommend you build vcl native apps and register multiple workers for better performance. Comments: - make sure sockets are not blocking otherwise there’s a good change that only one of the

[vpp-dev] lpm table api?

I see vpp uses a lpm table for fib_entry_t. Is any api available to create a lpm table in a plugin? Hemant smime.p7s Description: S/MIME cryptographic signature -=-=-=-=-=-=-=-=-=-=-=- Links: You receive all messages sent to this group. View/Reply Online (#21858):

Re: [vpp-dev] opensuse uplift

Hi, The correct link to FDio gerrit: https://gerrit.fd.io/r/c/vpp/+/37067 -- Laszlo Kiraly From: Laszlo Király Sent: Tuesday, September 6, 2022 6:50 PM To: vpp-dev@lists.fd.io Subject: opensuse uplift Hello devs,

Re: [vpp-dev] opensuse uplift

> It's just a notice, that I reworked a minor fix for build on openSUSE. > Excuse me if I confused the reviewer with my updates on this commit. > The link for gerrit review: https://gerrit.nordix.org/c/fdio/vpp/+/14826 Hi Laszlo, Thanks but the link points to gerrit.nordix.org which is not fd.io

[vpp-dev] UDP multithreading error

The attachment is my server program. My VPP version is 22.06. When I start this server by LDP and send a request to this server, the server will have a direct segment error. Two threads share an event queue, resulting in an event queue crash. Can you tell me if LDP does not support this use? In

[vpp-dev] opensuse uplift

Hello devs, It's just a notice, that I reworked a minor fix for build on openSUSE. Excuse me if I confused the reviewer with my updates on this commit. The link for gerrit review: https://gerrit.nordix.org/c/fdio/vpp/+/14826 -- Laszlo Kiraly

Re: [vpp-dev] One proposal for community meeting

Hi Xu, Presenting your patch is a fine idea. I will also spend a few minutes discussing the upgrade of the CI infra from ubuntu-20.04 to ubuntu-22.04. You have the correct link to join the meeting. Thanks, -daw- On 9/5/22 11:18 PM, Xu, Ting wrote: Hi, As I see in Dave’s message, our

<    4   5   6   7   8   9   10   11   12   13   >