Hi Ravi, At this time, there’s no option to pin connected udp sessions to workers, instead we default to worker 1. You could try changing the logic here [1] to select a random thread for each connection. We’re not doing that because whenever we receive return traffic we pin the session to the worker selected by rss to be sure the same worker is used for rx and tx.
I was asking about the max throughput you want to achieve because one udp session with 1.5kB datagrams should be able to go above 20Gbps. At least that’s what I’m seeing in my testbed with ldp + iperf3 -l 1460 -b 40g. Regards, Florin [1] https://git.fd.io/vpp/tree/src/vnet/udp/udp.c#n377 > On Sep 25, 2020, at 9:46 AM, RaviKiran Veldanda <[email protected]> wrote: > > Thanks Florin for your response > Please find my answers below > 1> I opened the UDP sessions with vppcom_session_connect and these are always > for TX and it will not receive any data. > 2> No there is no sent traffic on these. I can try this part. > 3> Actually the throughput as you know depends on different factors like > cores and etc, we have 100G card but we have our own application has limit. > So i cannot answer what maximum we can push. > 4> Coming to the dropped the Input Queues are getting filled with traffic. > Connection State Rx-f Tx-f [1:0][U] > 2001:5b0:ffff:1150:b883:31f:178:98f8:2035OPENED 0 7999863 [1:1][U] > 2001:5b0:ffff:1150:b883:31f:178:98f8:5100OPENED 0 7999863 [1:2][U] > 2001:5b0:ffff:1150:b883:31f:178:98f8:1160OPENED 0 7999863 [1:3][U] > 2001:5b0:ffff:1150:b883:31f:178:98f8:3442OPENED 0 7999863 [1:4][U] > 2001:5b0:ffff:1150:b883:31f:178:98f8:1220OPENED 0 7999863 [1:5][U] > 2001:5b0:ffff:1150:b883:31f:178:98f8:3303OPENED 0 7999863 [1:6][U] > 2001:5b0:ffff:1150:b883:31f:178:98f8:5397OPENED 0 0 [1:7][U] > 2001:5b0:ffff:1150:b883:31f:178:98f8:2545OPENED 0 0 [1:8][U] > 2001:5b0:ffff:1150:b883:31f:178:98f8:4097OPENED 0 0 [1:9][U] > 2001:5b0:ffff:1150:b883:31f:178:98f8:3618OPENED 0 0 [1:10][U] > 2001:5b0:ffff:1150:b883:31f:178:98f8:141OPENED 0 0 [1:11][U] > 2001:5b0:ffff:1150:b883:31f:178:98f8:465OPENED 0 0 Thread 1: active sessions > 12 > > The Interface Counters doesn't have drops, RX FIFO and TX-FIFO > rx-fifo-size 8000000 tx-fifo-size 8000000 > > Please let me know is there any other method to distribute these TX Flows or > increase the throughput. > //Ravi > > >
-=-=-=-=-=-=-=-=-=-=-=- Links: You receive all messages sent to this group. View/Reply Online (#17519): https://lists.fd.io/g/vpp-dev/message/17519 Mute This Topic: https://lists.fd.io/mt/77078709/21656 Mute #vnet:https://lists.fd.io/g/vpp-dev/mutehashtag/vnet Mute #vppcom:https://lists.fd.io/g/vpp-dev/mutehashtag/vppcom Mute #txbuff:https://lists.fd.io/g/vpp-dev/mutehashtag/txbuff Group Owner: [email protected] Unsubscribe: https://lists.fd.io/g/vpp-dev/unsub [[email protected]] -=-=-=-=-=-=-=-=-=-=-=-
