Hello all. I am attempting to run VPP in LXD managed containers with memif interfaces to the host VPP. However I am finding that the VPP instances in the containers "own" a CPU core apiece, which I thought should happen only with the host VPP due to the use of PMD drivers. Is there any way to avoid devoting a CPU core to each containerized VPP? That happens when the memif interface goes live, does it mean memif is like a PMD device?
In the event that I cannot use memif, giving the ratio of containers to available CPU resources, which is the lesser evil - tap or veth interfaces for the containers? And how much is that going to cost me in terms of performance and stability? Charles
-=-=-=-=-=-=-=-=-=-=-=- Links: You receive all messages sent to this group. View/Reply Online (#12270): https://lists.fd.io/g/vpp-dev/message/12270 Mute This Topic: https://lists.fd.io/mt/29886564/21656 Mute #vpp: https://lists.fd.io/mk?hashtag=vpp&subid=1480452 Group Owner: [email protected] Unsubscribe: https://lists.fd.io/g/vpp-dev/unsub [[email protected]] -=-=-=-=-=-=-=-=-=-=-=-
