No, Gulshan, my idea was a little bit different. I will try to explain it differently. For example you have 2 workers 0 and 1. You configure hardware NIC to RSS packets to both workers with some hash. Maybe even with the same hash. You process worker0 and detected some packet which will require long time processing. Some packets are already accepted and in RX ring. Because you already mapped NIC config space you can quickly reconfigure RSS settings to say NIC card to not send more packets to worker0. So RSS will send packets only to worker1. When worker0 finished process heavy load for that packet, it can process the rest of RX queue already accepted by the NIC. Then worker0 can turn on RSS for this worker0 and continue operate as it worked before. I did not do this practically but theoretically that is doable. Limitation in the system is number of workers. And it does not matter how you pass packets to additional workers: with queue or with hardware. I would say that hardware way might be more faster if it's doable. BR, Maxim. 04.02.2026, 06:38, [email protected] Uvarov if you want to remove those packets those are just easy task, you can just drop it on the card. There is a feature of filtering packets on Melanox. But i need those packets, i can not just simply remove them from queue. If that would have been the case then it would be a simple task.
-=-=-=-=-=-=-=-=-=-=-=- Links: You receive all messages sent to this group. View/Reply Online (#26782): https://lists.fd.io/g/vpp-dev/message/26782 Mute This Topic: https://lists.fd.io/mt/117523632/21656 Mute #vppctl:https://lists.fd.io/g/vpp-dev/mutehashtag/vppctl Mute #vpp-dev:https://lists.fd.io/g/vpp-dev/mutehashtag/vpp-dev Mute #vppinfra:https://lists.fd.io/g/vpp-dev/mutehashtag/vppinfra Mute #dpdk:https://lists.fd.io/g/vpp-dev/mutehashtag/dpdk Mute #mellanox:https://lists.fd.io/g/vpp-dev/mutehashtag/mellanox Group Owner: [email protected] Unsubscribe: https://lists.fd.io/g/vpp-dev/leave/14379924/21656/631435203/xyzzy [[email protected]] -=-=-=-=-=-=-=-=-=-=-=-
