[vpp-dev] fd problems when use cut-through connections

Hi there,  We are trying to use VPP cut-through feature (version 20.09).  What confuses me is why separate fifo segments  are used for each connection ?  In our case,  we  establish a lot of  cut-through connections,  and  we find that VPP send a lot of file descriptors to VCL  (

Re: [vpp-dev] fd problems when use cut-through connections

linux. Regards, FlorinOn Feb 25, 2021, at 11:51 PM, liuyacan <liuya...@corp.netease.com> wrote:Hi there,  We are trying to use VPP cut-through feature (version 20.09).  What confuses me is why separate fifo segments  are used for each connection ?  In our case,  we  establish a lot of  cut-t

Re: [vpp-dev] #vpp #vpp-dev

hI Nikhil, I don't know why the VM did not receive icmp reply packet, But I am sure it is not caused by fib because icmp reply, because the packet will not go through the ip4-lookup again.The reason you see "fib 0" is because the value of sw_if_index[VLIB_RX] was set to vnet_

[vpp-dev] TCP delay ACK

Hi All,     It seems that  macro TCP_ALWAYS_ACK is not used now，Are there any considerations? Is there any way I can Enable delay ACK ?Best regards,yacan -=-=-=-=-=-=-=-=-=-=-=- Links: You receive all messages sent to this group. View/Reply Online (#19004): https://lists.fd.io/g

Re: [vpp-dev] TCP delay ACK

rpose, the reasoning being that by construction we only send an ack per burst of packets received. Moreover, delayed acks tend to interfere with congestion control algorithms and rtt estimators. Out of curiosity, why are you looking into that?Regards,FlorinOn Mar 23, 2021, at 10:57 PM, liuyacan &

[vpp-dev] Is the tcp connection timestamp updated too early

Hi,     For ESTABLISHED tcp connection, its timestamp is updated in tcp_segment_validate()->tcp_options_parse(),     but if after this , the ACK field check failed,  then the packet  would be dropped.     Do we need to restore the timestamp of the connection ?Thanks,Yacan

Re: [vpp-dev] Is the tcp connection timestamp updated too early

On Mar 31, 2021, at 7:00 PM, liuyacan <liuya...@corp.netease.com> wrote:Hi,     For ESTABLISHED tcp connection, its timestamp is updated in tcp_segment_validate()->tcp_options_parse(),     but if after this , the ACK field check failed,  then the packet  would be dropped.     Do we need

Re: [vpp-dev] Is the tcp connection timestamp updated too early

know if you hit any other issues as I’ve only done Codenomicon testing. Regards,Florin[1] https://tools.ietf.org/html/rfc7323#appendix-DOn Mar 31, 2021, at 8:45 PM, liuyacan <liuya...@corp.netease.com> wrote:Hi Florin,   when we use packetdrill to inject a bad packet with high tsval whi

[vpp-dev] About cnat FIB

Hi,         We observed that CNAT translation rule now can only be set in the default table for v4 and v6.     Why is there such a restriction ? Best Regards,Yacan -=-=-=-=-=-=-=-=-=-=-=- Links: You receive all messages sent to th

Re: [vpp-dev] About cnat FIB

Maybe DNATing to a differenttable ?Best regards,-NathanLe mar. 6 avr. 2021 à 11:23, liuyacan <liuya...@corp.netease.com> a écrit : Hi,         We observed that CNAT translation rule now can only be set in the default table for v4 and v6.     Why is there such a restric

[vpp-dev] Can tcp4 packet bypass ip4-lookup node ?

Hi VPP experts,     The command "show node tcp4-output" tell me that  ip4-lookup and ip4-rewrite      are both possible  nodes for  Tx  packets. But I never saw the count of ip4-rewrite      greater than zero      The function tcp_output_handle_packet()  checks  tc0's next_no

Re: [vpp-dev] About cnat FIB

im at supporting all kube-proxy use-cases, you might require some additional evolutions.Kind regards,-Nathan[0] https://github.com/projectcalico/vpp-dataplaneLe mer. 7 avr. 2021 à 04:40, liuyacan <liuya...@corp.netease.com> a écrit : Hi  Nathan,   We hope to use CNAT

[vpp-dev] A question about 'dpo_get_next_node()'

Hi, All  I'm confused with the function 'dpo_get_next_node()'.   The comment says it create a graph arc from each of   the child's registered node types to "each" of the parent's.  But if the parent_indices is a vector, assertions will be   triggered. So it implies that calle

[vpp-dev] Fail to build vpp locally

Hi, All       I checked out the latest code and try to build, but encountered the following error.  Do I need to install/update any package ?I'm sure it was ok before. VPP version         : 21.06-rc0~755-g785458895VPP library version : 21.06GIT toplevel dir    : /home/liu

Re: [vpp-dev] test performance of nginx using vpp host stack#vpp-hoststack

Hi Florin, weizhen9612:     I'm not sure whether rpc for connects will be executed immediately by the main thread in the current implementation, or will it wait for the epoll_pwait in linux_epoll_input_inline() to time out. Regards,yacan On 5/5/2022 16:19， wro

Re: [vpp-dev] test performance of nginx using vpp host stack#vpp-hoststack

improvements are needed.   Regards,FlorinOn May 5, 2022, at 5:01 AM, liuyacan <liuya...@corp.netease.com> wrote:Hi Florin, weizhen9612:     I'm not sure whether rpc for connects will be executed immediately by the main thread in the current implementation, or will it wait for the epo