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 VPP). So if traffic is forwarded from/to this interfaces only, it should be using 8 cores and you should reach performance much higher than 14Mpps. The tap interface is created by lcp, and as far as I understand it, it is used as a "representor" for Linux but it should not be used for most traffic apart maybe some control-plane traffic. The fact that is uses only 1rxq (and hence 1 core) shouldn't be an issue. If you want us to help on performance issue, please share the output of: ~# vppctl clear run && sleep 1 && vppctl show run While the traffic is going through? That should show us where CPU cycles are spent and how traffic distribution looks like.
Best ben > -----Original Message----- > From: vpp-dev@lists.fd.io <vpp-dev@lists.fd.io> On Behalf Of Pawel > Staszewski > Sent: Thursday, September 15, 2022 0:03 > To: vpp-dev@lists.fd.io > Subject: 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 interface:Interface: tap3 (ifindex 5) > name "e0-rdma" > host-mtu-size "9000" > host-mac-addr: 02:fe:ad:1c:47:da > host-carrier-up: 1 > vhost-fds 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 > tap-fds 58 > gso-enabled 0 > csum-enabled 0 > packet-coalesce 0 > packet-buffering 0 > Mac Address: 02:fe:57:90:b6:d7 > Device instance: 1 > flags 0x1 > admin-up (0) > features 0x110008000 > VIRTIO_NET_F_MRG_RXBUF (15) > VIRTIO_RING_F_INDIRECT_DESC (28) > VIRTIO_F_VERSION_1 (32) > remote-features 0x33d008000 > VIRTIO_NET_F_MRG_RXBUF (15) > VIRTIO_F_NOTIFY_ON_EMPTY (24) > VHOST_F_LOG_ALL (26) > VIRTIO_F_ANY_LAYOUT (27) > VIRTIO_RING_F_INDIRECT_DESC (28) > VIRTIO_RING_F_EVENT_IDX (29) > VIRTIO_F_VERSION_1 (32) > VIRTIO_F_IOMMU_PLATFORM (33) > Number of RX Virtqueue 1 > Number of TX Virtqueue 17 > > > > > > > hmm > > Number of RX Virtqueue 1 > > > Number of TX Virtqueue 17 > > > > > Why ? > > > > > When > > vpp# show hardware-interfaces enp59s0f0-rdma > Name Idx Link Hardware > enp59s0f0-rdma 3 up enp59s0f0-rdma > Link speed: 100 Gbps > RX Queues: > queue thread mode > 0 vpp_wk_1 (2) polling > 1 vpp_wk_2 (3) polling > 2 vpp_wk_3 (4) polling > 3 vpp_wk_4 (5) polling > 4 vpp_wk_5 (6) polling > 5 vpp_wk_6 (7) polling > 6 vpp_wk_7 (8) polling > 7 vpp_wk_8 (9) polling > Ethernet address 0c:42:a1:a9:6c:a6 > netdev enp59s0f0 pci-addr 0000:3b:00.0 > product name: CX516A - ConnectX-5 QSFP28 > part number: MCX516A-CCAT > revision: AC > serial number: MT2024K11216 > flags: admin-up link-up mlx5dv > rss: ipv4-tcp ipv6-tcp > mlx5: version 0 > device flags: cqe-v1 mpw-allowed enhanced-mpw cqe-128b-comp cqe-128b- > pad > vpp# > > > > > > > > remember this is not dpdk - but I need to create rdma interfaces first > > But it looks like there is some problems with tap interface queue assign > > > > > > > > > > > > On 9/12/22 12:20 PM, Benoit Ganne (bganne) via lists.fd.io wrote: > > > This is what we measure on CSIT with Intel Cascade Lake for 1 CPU > core: > http://csit.fd.io/trending/#eNrVlMFuwyAMhp8mvVSWwC3JduihXd6jYuAu0Uj > KgFbtnn6kyuRE2ibtUm0HjNBn-8e_EDEdA- > 0juU2hdkW1K7BqbQ7FarvM29l7wB6Mu4AU4oXQS3NRqtQQbKeh9Wso188gDVBqhlNe0WhHKF4h > 9Ba6EIZu- > DR0s6c0a83EN1cm3wtygQ6kuSKrMkoUJzJf34uzD0F3FNt34pI8EXOT_WEkzVwnXf2EjsNW9S3 > jt2Y6nHiZT14n09zJQIc_- > Pd5l79iGj1IYd5Anw_3eYCs92_en6oXMemQNigQQaxAlkshctoYFtTbkT2CxDnrj6G7fQeq_gA > VCBnd > - red line is Mellanox Cx5 l2 patch (ie max perf we can sustain > for this HW setup) and yields ~42Mpps > - back line is Mellanox Cx5 IPv4 forwarding with 20k prefixes and > random traffic (should be close to your case I suppose) and yields ~15Mpps > - blue line is Intel Columbiaville IPv4 forwarding with 20k > prefixes and random traffic (for reference) and yields ~18Mpps > > Your results seem to be close to what we see *for a single core*. > So now I'm wondering, is your traffic distributed on all the cores? > To go further, can you share the output of: > ~# vppctl clear run && sleep 1 && vppctl show run > While the traffic is going through? That should show us where CPU > cycles are spent and how traffic distribution looks like. > > Best > ben > > > -----Original Message----- > From: vpp-dev@lists.fd.io <mailto:vpp-dev@lists.fd.io> > <vpp-dev@lists.fd.io> <mailto:vpp-dev@lists.fd.io> On Behalf Of Pawel > Staszewski > Sent: Sunday, September 11, 2022 22:34 > To: vpp-dev@lists.fd.io <mailto:vpp-dev@lists.fd.io> > Subject: [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 RX queue > - and vpp config has cpu configuration where we are using > 25(vpp only) > cores for all workers (controll+24 workers and 8 cores left > for frr > software routes and other stuff ) > > We are using isolcpus where last 24 cores are isolated with > nohz nocb > etc... > > > Hardware used is : > > Intel 6246R > > 96GB ram > > Mellanox Connect-X 5 2x 100GB Ethernet NIC > > > -------- > > And the problem is that we can reach only like 7Mpps per > port (2 ports > are bonded) - so in total 14Mpps for two ports in bonding > > Bonding is done via vppctl > > bond is reporting drops in stats - but this stats shows "no > error" > > Example drops from bond: > > vpp# show interface > Name Idx State MTU > (L3/IP4/IP6/MPLS) Counter Count > BondEthernet0 5 up 1500/0/0/0 > rx > packets 16140265734 > > rx > bytes 14149099692091 > > tx > packets 16136936492 > > tx > bytes 14148733722122 > drops 3451254 > punt 1326 > tx-error 1 > > > And show errors: > > vpp# show errors > Count Node Reason > Severity > 16 ip6-icmp-input neighbor > solicitations for > unknown error > 952172 bond-input > no > error error > 1 null-node > blackholed > packets error > 2 lacp-input good lacp > packets -- > cache hit error > 24 arp-reply ARP > replies > sent info > 9 ip4-input > ip4 ttl <= > 1 error > 32 ip6-icmp-input neighbor > solicitations for > unknown error > 1 ip6-icmp-input neighbor > advertisements sent info > 9 ip4-icmp-error hop limit > exceeded > response sent info > 1772997 bond-input > no > error error > 2 bond-input pass through > (CDP, LLDP, > slow proto error > 140 snap-input unknown > oui/snap > protocol error > 180 ethernet-input > no > error error > 1 ethernet-input unknown > ethernet > type error > 61 ethernet-input > unknown > vlan error > 1 null-node > blackholed > packets error > 12 ip4-input > ip4 ttl <= > 1 error > 1 ip6-icmp-input neighbor > solicitations > from source error > 30 ip6-icmp-input neighbor > solicitations for > unknown error > 12 ip4-icmp-error hop limit > exceeded > response sent info > 1853434 bond-input > no > error error > 1 ethernet-input > unknown > vlan error > 1 ip4-input > ip4 ttl <= > 1 error > 1 ip4-local ip4 > source lookup > miss error > 30 ip6-icmp-input neighbor > solicitations for > unknown error > 1 ip6-icmp-input neighbor > advertisements sent info > 1 ip4-icmp-error hop limit > exceeded > response sent info > 1864191 bond-input > no > error error > 1 null-node > blackholed > packets error > 5 ip4-input > ip4 ttl <= > 1 error > 35 ip6-icmp-input neighbor > solicitations for > unknown error > 5 ip4-icmp-error hop limit > exceeded > response sent info > 1743015 bond-input > no > error error > 3 ethernet-input > unknown > vlan error > 1 null-node > blackholed > packets error > 4 ip4-input > ip4 ttl <= > 1 error > 1 ip6-icmp-input neighbor > solicitations > from source error > 35 ip6-icmp-input neighbor > solicitations for > unknown error > 4 ip4-icmp-error hop limit > exceeded > response sent info > 1917745 bond-input > no > error error > 2 ip4-input > ip4 ttl <= > 1 error > 18 ip6-icmp-input neighbor > solicitations for > unknown error > 2 ip4-icmp-error hop limit > exceeded > response sent info > 886249 bond-input > no > error error > 1 ethernet-input > unknown > vlan error > 7 ip4-input > ip4 ttl <= > 1 error > 1 ip6-icmp-input > valid > packets info > 14 ip6-icmp-input neighbor > solicitations for > unknown error > 1 ip6-icmp-input router > advertisements > received info > > > This counter : > > 1743015 bond-input > no > error error > > Is wild :) and it increasing really fast - and also simple > icmp ping to > this host or from this host is reporting froms from 2 to 5% > - when both > 100Gnics are receiving and transmitting 7Mpps - per port ( > so in total > 14Mpps RX and 14Mpps TX) > > Basically it looks like something cant handle more - and > probabbly > FIB/RIB processing is the case here - because without > forwarding and > frr+routes we was able to reach like 40Mpps per 100G port > > > Bonded nics: (m0rdma0+m1rdma1 - created by lcp from two > physical mlx > ports as RDMA pairs) - create commands: > > create int rdma host-if ens1f1np1 name m1rdma1 num-rx-queues > 12 > create int rdma host-if ens1f0np0 name m0rdma0 num-rx-queues > 12 > > create bond mode lacp load-balance l34 > bond add BondEthernet0 m0rdma0 > bond add BondEthernet0 m1rdma1 > > > Nics attached via rdma: > > m0rdma0 4 up m0rdma0 > Link speed: 100 Gbps > RX Queues: > queue thread mode > 0 vpp_wk_13 (14) polling > 1 vpp_wk_14 (15) polling > 2 vpp_wk_15 (16) polling > 3 vpp_wk_16 (17) polling > 4 vpp_wk_17 (18) polling > 5 vpp_wk_18 (19) polling > 6 vpp_wk_19 (20) polling > 7 vpp_wk_20 (21) polling > 8 vpp_wk_21 (22) polling > 9 vpp_wk_22 (23) polling > 10 vpp_wk_23 (24) polling > 11 vpp_wk_0 (1) polling > Ethernet address 98:03:9b:67:f6:1e > netdev ens1f0np0 pci-addr 0000:b3:00.0 > product name: CX516A - ConnectX-5 QSFP28 > part number: MCX516A-CCAT > revision: B2 > serial number: MT2043K02404 > flags: admin-up mlx5dv > rss: ipv4-tcp ipv6-tcp > mlx5: version 0 > device flags: cqe-v1 mpw-allowed enhanced-mpw cqe- > 128b-comp > cqe-128b-pad > m1rdma1 3 up m1rdma1 > Link speed: 100 Gbps > RX Queues: > queue thread mode > 0 vpp_wk_1 (2) polling > 1 vpp_wk_2 (3) polling > 2 vpp_wk_3 (4) polling > 3 vpp_wk_4 (5) polling > 4 vpp_wk_5 (6) polling > 5 vpp_wk_6 (7) polling > 6 vpp_wk_7 (8) polling > 7 vpp_wk_8 (9) polling > 8 vpp_wk_9 (10) polling > 9 vpp_wk_10 (11) polling > 10 vpp_wk_11 (12) polling > 11 vpp_wk_12 (13) polling > Ethernet address 98:03:9b:67:f6:1e > netdev ens1f1np1 pci-addr 0000:b3:00.1 > product name: CX516A - ConnectX-5 QSFP28 > part number: MCX516A-CCAT > revision: B2 > serial number: MT2043K02404 > flags: admin-up mlx5dv > rss: ipv4-tcp ipv6-tcp > mlx5: version 0 > device flags: cqe-v1 mpw-allowed enhanced-mpw cqe- > 128b-comp > cqe-128b-pad > > .... > > .... > > Same NIC same Queue threads numbers (12 per card from 24 > available cores > that are isolcpus - at kernel boot) > > > vpp# show interface rx-placement > Thread 1 (vpp_wk_0): > node rdma-input: > m0rdma0 queue 11 (polling) > node virtio-input: > tap1 queue 0 (polling) > Thread 2 (vpp_wk_1): > node rdma-input: > m1rdma1 queue 0 (polling) > node virtio-input: > tap5 queue 0 (polling) > Thread 3 (vpp_wk_2): > node rdma-input: > m1rdma1 queue 1 (polling) > Thread 4 (vpp_wk_3): > m1rdma1 queue 2 (polling) > Thread 5 (vpp_wk_4): > m1rdma1 queue 3 (polling) > Thread 6 (vpp_wk_5): > m1rdma1 queue 4 (polling) > Thread 7 (vpp_wk_6): > m1rdma1 queue 5 (polling) > Thread 8 (vpp_wk_7): > m1rdma1 queue 6 (polling) > Thread 9 (vpp_wk_8): > m1rdma1 queue 7 (polling) > Thread 10 (vpp_wk_9): > m1rdma1 queue 8 (polling) > Thread 11 (vpp_wk_10): > m1rdma1 queue 9 (polling) > Thread 12 (vpp_wk_11): > m1rdma1 queue 10 (polling) > Thread 13 (vpp_wk_12): > m1rdma1 queue 11 (polling) > Thread 14 (vpp_wk_13): > m0rdma0 queue 0 (polling) > Thread 15 (vpp_wk_14): > m0rdma0 queue 1 (polling) > Thread 16 (vpp_wk_15): > m0rdma0 queue 2 (polling) > Thread 17 (vpp_wk_16): > m0rdma0 queue 3 (polling) > Thread 18 (vpp_wk_17): > m0rdma0 queue 4 (polling) > Thread 19 (vpp_wk_18): > m0rdma0 queue 5 (polling) > Thread 20 (vpp_wk_19): > m0rdma0 queue 6 (polling) > Thread 21 (vpp_wk_20): > m0rdma0 queue 7 (polling) > Thread 22 (vpp_wk_21): > m0rdma0 queue 8 (polling) > Thread 23 (vpp_wk_22): > m0rdma0 queue 9 (polling) > Thread 24 (vpp_wk_23): > m0rdma0 queue 10 (polling) > > Below also some stats about memory used: > > vpp# show memory > show memory: Need one of api-segment, stats-segment, main- > heap, > numa-heaps or map > vpp# show memory main-heap > Thread 0 vpp_main > base 0x7f64c0000000, size 2g, locked, unmap-on-destroy, > name 'main > heap' > page stats: page-size 1G, total 2, mapped 2, not-mapped > 0 > numa 0: 2 pages, 2g bytes > total: 1.99G, used: 369.37M, free: 1.64G, trimmable: > 1.62G > > Thread 1 vpp_wk_0 > base 0x7f64c0000000, size 2g, locked, unmap-on-destroy, > name 'main > heap' > page stats: page-size 1G, total 2, mapped 2, not-mapped > 0 > numa 0: 2 pages, 2g bytes > total: 1.99G, used: 369.37M, free: 1.64G, trimmable: > 1.62G > > Thread 2 vpp_wk_1 > base 0x7f64c0000000, size 2g, locked, unmap-on-destroy, > name 'main > heap' > page stats: page-size 1G, total 2, mapped 2, not-mapped > 0 > numa 0: 2 pages, 2g bytes > total: 1.99G, used: 369.38M, free: 1.64G, trimmable: > 1.62G > > Thread 3 vpp_wk_2 > base 0x7f64c0000000, size 2g, locked, unmap-on-destroy, > name 'main > heap' > page stats: page-size 1G, total 2, mapped 2, not-mapped > 0 > numa 0: 2 pages, 2g bytes > total: 1.99G, used: 369.38M, free: 1.64G, trimmable: > 1.62G > > Thread 4 vpp_wk_3 > base 0x7f64c0000000, size 2g, locked, unmap-on-destroy, > name 'main > heap' > page stats: page-size 1G, total 2, mapped 2, not-mapped > 0 > numa 0: 2 pages, 2g bytes > total: 1.99G, used: 369.38M, free: 1.64G, trimmable: > 1.62G > > Thread 5 vpp_wk_4 > base 0x7f64c0000000, size 2g, locked, unmap-on-destroy, > name 'main > heap' > page stats: page-size 1G, total 2, mapped 2, not-mapped > 0 > numa 0: 2 pages, 2g bytes > total: 1.99G, used: 369.38M, free: 1.64G, trimmable: > 1.62G > > Thread 6 vpp_wk_5 > base 0x7f64c0000000, size 2g, locked, unmap-on-destroy, > name 'main > heap' > page stats: page-size 1G, total 2, mapped 2, not-mapped > 0 > numa 0: 2 pages, 2g bytes > total: 1.99G, used: 369.38M, free: 1.64G, trimmable: > 1.62G > > Thread 7 vpp_wk_6 > base 0x7f64c0000000, size 2g, locked, unmap-on-destroy, > name 'main > heap' > page stats: page-size 1G, total 2, mapped 2, not-mapped > 0 > numa 0: 2 pages, 2g bytes > total: 1.99G, used: 369.38M, free: 1.64G, trimmable: > 1.62G > > Thread 8 vpp_wk_7 > base 0x7f64c0000000, size 2g, locked, unmap-on-destroy, > name 'main > heap' > page stats: page-size 1G, total 2, mapped 2, not-mapped > 0 > numa 0: 2 pages, 2g bytes > total: 1.99G, used: 369.38M, free: 1.64G, trimmable: > 1.62G > > Thread 9 vpp_wk_8 > base 0x7f64c0000000, size 2g, locked, unmap-on-destroy, > name 'main > heap' > page stats: page-size 1G, total 2, mapped 2, not-mapped > 0 > numa 0: 2 pages, 2g bytes > total: 1.99G, used: 369.38M, free: 1.64G, trimmable: > 1.62G > > Thread 10 vpp_wk_9 > base 0x7f64c0000000, size 2g, locked, unmap-on-destroy, > name 'main > heap' > page stats: page-size 1G, total 2, mapped 2, not-mapped > 0 > numa 0: 2 pages, 2g bytes > total: 1.99G, used: 369.38M, free: 1.64G, trimmable: > 1.62G > > Thread 11 vpp_wk_10 > base 0x7f64c0000000, size 2g, locked, unmap-on-destroy, > name 'main > heap' > page stats: page-size 1G, total 2, mapped 2, not-mapped > 0 > numa 0: 2 pages, 2g bytes > total: 1.99G, used: 369.38M, free: 1.64G, trimmable: > 1.62G > > Thread 12 vpp_wk_11 > base 0x7f64c0000000, size 2g, locked, unmap-on-destroy, > name 'main > heap' > page stats: page-size 1G, total 2, mapped 2, not-mapped > 0 > numa 0: 2 pages, 2g bytes > total: 1.99G, used: 369.38M, free: 1.64G, trimmable: > 1.62G > > Thread 13 vpp_wk_12 > base 0x7f64c0000000, size 2g, locked, unmap-on-destroy, > name 'main > heap' > page stats: page-size 1G, total 2, mapped 2, not-mapped > 0 > numa 0: 2 pages, 2g bytes > total: 1.99G, used: 369.38M, free: 1.64G, trimmable: > 1.62G > > Thread 14 vpp_wk_13 > base 0x7f64c0000000, size 2g, locked, unmap-on-destroy, > name 'main > heap' > page stats: page-size 1G, total 2, mapped 2, not-mapped > 0 > numa 0: 2 pages, 2g bytes > total: 1.99G, used: 369.38M, free: 1.64G, trimmable: > 1.62G > > Thread 15 vpp_wk_14 > base 0x7f64c0000000, size 2g, locked, unmap-on-destroy, > name 'main > heap' > page stats: page-size 1G, total 2, mapped 2, not-mapped > 0 > numa 0: 2 pages, 2g bytes > total: 1.99G, used: 369.38M, free: 1.64G, trimmable: > 1.62G > > Thread 16 vpp_wk_15 > base 0x7f64c0000000, size 2g, locked, unmap-on-destroy, > name 'main > heap' > page stats: page-size 1G, total 2, mapped 2, not-mapped > 0 > numa 0: 2 pages, 2g bytes > total: 1.99G, used: 369.38M, free: 1.64G, trimmable: > 1.62G > > Thread 17 vpp_wk_16 > base 0x7f64c0000000, size 2g, locked, unmap-on-destroy, > name 'main > heap' > page stats: page-size 1G, total 2, mapped 2, not-mapped > 0 > numa 0: 2 pages, 2g bytes > total: 1.99G, used: 369.38M, free: 1.64G, trimmable: > 1.62G > > Thread 18 vpp_wk_17 > base 0x7f64c0000000, size 2g, locked, unmap-on-destroy, > name 'main > heap' > page stats: page-size 1G, total 2, mapped 2, not-mapped > 0 > numa 0: 2 pages, 2g bytes > total: 1.99G, used: 369.38M, free: 1.64G, trimmable: > 1.62G > > Thread 19 vpp_wk_18 > base 0x7f64c0000000, size 2g, locked, unmap-on-destroy, > name 'main > heap' > page stats: page-size 1G, total 2, mapped 2, not-mapped > 0 > numa 0: 2 pages, 2g bytes > total: 1.99G, used: 369.38M, free: 1.64G, trimmable: > 1.62G > > Thread 20 vpp_wk_19 > base 0x7f64c0000000, size 2g, locked, unmap-on-destroy, > name 'main > heap' > page stats: page-size 1G, total 2, mapped 2, not-mapped > 0 > numa 0: 2 pages, 2g bytes > total: 1.99G, used: 369.38M, free: 1.64G, trimmable: > 1.62G > > Thread 21 vpp_wk_20 > base 0x7f64c0000000, size 2g, locked, unmap-on-destroy, > name 'main > heap' > page stats: page-size 1G, total 2, mapped 2, not-mapped > 0 > numa 0: 2 pages, 2g bytes > total: 1.99G, used: 369.38M, free: 1.64G, trimmable: > 1.62G > > Thread 22 vpp_wk_21 > base 0x7f64c0000000, size 2g, locked, unmap-on-destroy, > name 'main > heap' > page stats: page-size 1G, total 2, mapped 2, not-mapped > 0 > numa 0: 2 pages, 2g bytes > total: 1.99G, used: 369.38M, free: 1.64G, trimmable: > 1.62G > > Thread 23 vpp_wk_22 > base 0x7f64c0000000, size 2g, locked, unmap-on-destroy, > name 'main > heap' > page stats: page-size 1G, total 2, mapped 2, not-mapped > 0 > numa 0: 2 pages, 2g bytes > total: 1.99G, used: 369.38M, free: 1.64G, trimmable: > 1.62G > > Thread 24 vpp_wk_23 > base 0x7f64c0000000, size 2g, locked, unmap-on-destroy, > name 'main > heap' > page stats: page-size 1G, total 2, mapped 2, not-mapped > 0 > numa 0: 2 pages, 2g bytes > total: 1.99G, used: 369.38M, free: 1.64G, trimmable: > 1.62G > > Thread 25 vpp_wk_24 > base 0x7f64c0000000, size 2g, locked, unmap-on-destroy, > name 'main > heap' > page stats: page-size 1G, total 2, mapped 2, not-mapped > 0 > numa 0: 2 pages, 2g bytes > total: 1.99G, used: 369.38M, free: 1.64G, trimmable: > 1.62G > > Thread 26 vpp_wk_25 > base 0x7f64c0000000, size 2g, locked, unmap-on-destroy, > name 'main > heap' > page stats: page-size 1G, total 2, mapped 2, not-mapped > 0 > numa 0: 2 pages, 2g bytes > total: 1.99G, used: 369.38M, free: 1.64G, trimmable: > 1.62G > > Thread 27 vpp_wk_26 > base 0x7f64c0000000, size 2g, locked, unmap-on-destroy, > name 'main > heap' > page stats: page-size 1G, total 2, mapped 2, not-mapped > 0 > numa 0: 2 pages, 2g bytes > total: 1.99G, used: 369.39M, free: 1.64G, trimmable: > 1.62G > vpp# show memory main-heap > Thread 0 vpp_main > base 0x7f64c0000000, size 2g, locked, unmap-on-destroy, > name 'main > heap' > page stats: page-size 1G, total 2, mapped 2, not-mapped > 0 > numa 0: 2 pages, 2g bytes > total: 1.99G, used: 369.58M, free: 1.64G, trimmable: > 1.62G > > Thread 1 vpp_wk_0 > base 0x7f64c0000000, size 2g, locked, unmap-on-destroy, > name 'main > heap' > page stats: page-size 1G, total 2, mapped 2, not-mapped > 0 > numa 0: 2 pages, 2g bytes > total: 1.99G, used: 369.58M, free: 1.64G, trimmable: > 1.62G > > Thread 2 vpp_wk_1 > base 0x7f64c0000000, size 2g, locked, unmap-on-destroy, > name 'main > heap' > page stats: page-size 1G, total 2, mapped 2, not-mapped > 0 > numa 0: 2 pages, 2g bytes > total: 1.99G, used: 369.58M, free: 1.64G, trimmable: > 1.62G > > Thread 3 vpp_wk_2 > base 0x7f64c0000000, size 2g, locked, unmap-on-destroy, > name 'main > heap' > page stats: page-size 1G, total 2, mapped 2, not-mapped > 0 > numa 0: 2 pages, 2g bytes > total: 1.99G, used: 369.58M, free: 1.64G, trimmable: > 1.62G > > Thread 4 vpp_wk_3 > base 0x7f64c0000000, size 2g, locked, unmap-on-destroy, > name 'main > heap' > page stats: page-size 1G, total 2, mapped 2, not-mapped > 0 > numa 0: 2 pages, 2g bytes > total: 1.99G, used: 369.58M, free: 1.64G, trimmable: > 1.62G > > Thread 5 vpp_wk_4 > base 0x7f64c0000000, size 2g, locked, unmap-on-destroy, > name 'main > heap' > page stats: page-size 1G, total 2, mapped 2, not-mapped > 0 > numa 0: 2 pages, 2g bytes > total: 1.99G, used: 369.58M, free: 1.64G, trimmable: > 1.62G > > Thread 6 vpp_wk_5 > base 0x7f64c0000000, size 2g, locked, unmap-on-destroy, > name 'main > heap' > page stats: page-size 1G, total 2, mapped 2, not-mapped > 0 > numa 0: 2 pages, 2g bytes > total: 1.99G, used: 369.58M, free: 1.64G, trimmable: > 1.62G > > Thread 7 vpp_wk_6 > base 0x7f64c0000000, size 2g, locked, unmap-on-destroy, > name 'main > heap' > page stats: page-size 1G, total 2, mapped 2, not-mapped > 0 > numa 0: 2 pages, 2g bytes > total: 1.99G, used: 369.58M, free: 1.64G, trimmable: > 1.62G > > Thread 8 vpp_wk_7 > base 0x7f64c0000000, size 2g, locked, unmap-on-destroy, > name 'main > heap' > vpp# show memory main-heap > Thread 0 vpp_main > base 0x7f64c0000000, size 2g, locked, unmap-on-destroy, > name 'main > heap' > page stats: page-size 1G, total 2, mapped 2, not-mapped > 0 > numa 0: 2 pages, 2g bytes > total: 1.99G, used: 369.58M, free: 1.64G, trimmable: > 1.62G > > Thread 1 vpp_wk_0 > base 0x7f64c0000000, size 2g, locked, unmap-on-destroy, > name 'main > heap' > page stats: page-size 1G, total 2, mapped 2, not-mapped > 0 > numa 0: 2 pages, 2g bytes > total: 1.99G, used: 369.58M, free: 1.64G, trimmable: > 1.62G > > Thread 2 vpp_wk_1 > base 0x7f64c0000000, size 2g, locked, unmap-on-destroy, > name 'main > heap' > page stats: page-size 1G, total 2, mapped 2, not-mapped > 0 > numa 0: 2 pages, 2g bytes > total: 1.99G, used: 369.58M, free: 1.64G, trimmable: > 1.62G > > Thread 3 vpp_wk_2 > base 0x7f64c0000000, size 2g, locked, unmap-on-destroy, > name 'main > heap' > page stats: page-size 1G, total 2, mapped 2, not-mapped > 0 > numa 0: 2 pages, 2g bytes > total: 1.99G, used: 369.58M, free: 1.64G, trimmable: > 1.62G > > Thread 4 vpp_wk_3 > base 0x7f64c0000000, size 2g, locked, unmap-on-destroy, > name 'main > heap' > page stats: page-size 1G, total 2, mapped 2, not-mapped > 0 > numa 0: 2 pages, 2g bytes > total: 1.99G, used: 369.58M, free: 1.64G, trimmable: > 1.62G > > Thread 5 vpp_wk_4 > base 0x7f64c0000000, size 2g, locked, unmap-on-destroy, > name 'main > heap' > page stats: page-size 1G, total 2, mapped 2, not-mapped > 0 > numa 0: 2 pages, 2g bytes > total: 1.99G, used: 369.58M, free: 1.64G, trimmable: > 1.62G > > Thread 6 vpp_wk_5 > base 0x7f64c0000000, size 2g, locked, unmap-on-destroy, > name 'main > heap' > page stats: page-size 1G, total 2, mapped 2, not-mapped > 0 > numa 0: 2 pages, 2g bytes > total: 1.99G, used: 369.58M, free: 1.64G, trimmable: > 1.62G > > Thread 7 vpp_wk_6 > base 0x7f64c0000000, size 2g, locked, unmap-on-destroy, > name 'main > heap' > page stats: page-size 1G, total 2, mapped 2, not-mapped > 0 > numa 0: 2 pages, 2g bytes > total: 1.99G, used: 369.58M, free: 1.64G, trimmable: > 1.62G > > Thread 8 vpp_wk_7 > base 0x7f64c0000000, size 2g, locked, unmap-on-destroy, > name 'main > heap' > vpp# show memory ? > show memory show memory > [api-segment][stats-segment][verbose] > [numa-heaps][map] > vpp# show memory > show memory: Need one of api-segment, stats-segment, main- > heap, > numa-heaps or map > vpp# show memory numa-heaps > Numa 0 uses the main heap... > vpp# show memory map > > StartAddr size FD PageSz Pages Numa0 NotMap Name > 00007f64c0000000 2g 1G 2 2 0 main > heap > 00007f655b663000 2m 4K 512 9 503 > thread stack: > thread 0 > 00007f6440000000 1g 4 1G 1 1 0 stat > segment > 00007f655bc9a000 32k 4K 8 2 6 > process stack: > wg-timer-manager > 00007f655b642000 128k 4K 32 2 30 > process stack: > vrrp-periodic-process > 00007f655bc7a000 32k 4K 8 1 7 > process stack: > prom-scraper-process > 00007f655b639000 32k 4K 8 2 6 > process stack: > nsh-md2-ioam-export-process > 00007f655b630000 32k 4K 8 2 6 > process stack: > nat44-ei-ha-process > 00007f655b627000 32k 4K 8 2 6 > process stack: > memif-process > 00007f655b61e000 32k 4K 8 2 6 > process stack: > lldp-process > 00007f655b5fd000 128k 4K 32 3 29 > process stack: > linux-cp-netlink-process > 00007f655b5f4000 32k 4K 8 2 6 > process stack: > udp-ping-process > 00007f655b5eb000 32k 4K 8 2 6 > process stack: > vxlan-gpe-ioam-export-process > 00007f655b5e2000 32k 4K 8 2 6 > process stack: > ioam-export-process > 00007f655b5d9000 32k 4K 8 2 6 > process stack: > ikev2-manager-process > 00007f655b5d0000 32k 4K 8 2 6 > process stack: > igmp-timer-process > 00007f655b5c7000 32k 4K 8 1 7 > process stack: > http-timer-process > 00007f655b5be000 32k 4K 8 2 6 > process stack: > flowprobe-timer-process > 00007f655b59d000 128k 4K 32 2 30 > process stack: > dpdk-process > 00007f655b57c000 128k 4K 32 2 30 > process stack: > admin-up-down-process > 00007f655b573000 32k 4K 8 2 6 > process stack: > send-dhcp6-pd-client-message-process > 00007f655b56a000 32k 4K 8 2 6 > process stack: > dhcp6-pd-client-cp-process > 00007f655b561000 32k 4K 8 2 6 > process stack: > dhcp6-client-cp-process > 00007f655b558000 32k 4K 8 2 6 > process stack: > send-dhcp6-client-message-process > 00007f655b54f000 32k 4K 8 2 6 > process stack: > dhcp6-pd-reply-publisher-process > 00007f655b546000 32k 4K 8 2 6 > process stack: > dhcp6-reply-publisher-process > 00007f655b535000 64k 4K 16 2 14 > process stack: > dhcp-client-process > 00007f655b52c000 32k 4K 8 2 6 > process stack: > cnat-scanner-process > 00007f655b523000 32k 4K 8 2 6 > process stack: > avf-process > 00007f655b51a000 32k 4K 8 2 6 > process stack: > acl-plugin-fa-cleaner-process > 00007f655b4d9000 256k 4K 64 3 61 > process stack: > api-rx-from-ring > 00007f655b4d0000 32k 4K 8 2 6 > process stack: > rd-cp-process > 00007f655b4c7000 32k 4K 8 2 6 > process stack: > ip6-ra-process > 00007f655b4be000 32k 4K 8 2 6 > process stack: > ip6-rs-process > 00007f655b4b5000 32k 4K 8 2 6 > process stack: > ip6-mld-process > 00007f655b4ac000 32k 4K 8 2 6 > process stack: > fib-walk > 00007f655b4a3000 32k 4K 8 1 7 > process stack: > session-queue-process > 00007f655b49a000 32k 4K 8 2 6 > process stack: > vhost-user-process > 00007f655b491000 32k 4K 8 2 6 > process stack: > vhost-user-send-interrupt-process > 00007f655b488000 32k 4K 8 2 6 > process stack: > flow-report-process > 00007f655b47f000 32k 4K 8 2 6 > process stack: > bfd-process > 00007f655b476000 32k 4K 8 2 6 > process stack: > ip-neighbor-event > 00007f655b46d000 32k 4K 8 2 6 > process stack: > ip6-neighbor-age-process > 00007f655b464000 32k 4K 8 2 6 > process stack: > ip4-neighbor-age-process > 00007f655b45b000 32k 4K 8 2 6 > process stack: > ip6-sv-reassembly-expire-walk > 00007f655b452000 32k 4K 8 2 6 > process stack: > ip6-full-reassembly-expire-walk > 00007f655b449000 32k 4K 8 2 6 > process stack: > ip4-sv-reassembly-expire-walk > 00007f655b440000 32k 4K 8 2 6 > process stack: > ip4-full-reassembly-expire-walk > 00007f655b437000 32k 4K 8 2 6 > process stack: > bond-process > 00007f655b42e000 32k 4K 8 2 6 > process stack: > l2fib-mac-age-scanner-process > 00007f655b425000 32k 4K 8 2 6 > process stack: > l2-arp-term-publisher > 00007f655b41c000 32k 4K 8 2 6 > process stack: > vpe-link-state-process > 00007f655b3db000 256k 4K 64 5 59 > process stack: > startup-config-process > 00007f655b3d2000 32k 4K 8 2 6 > process stack: > statseg-collector-process > 00007f655a5a8000 2m 4K 512 7 505 > thread stack: > thread 1 > 00007f655a3a7000 2m 4K 512 10 502 > thread stack: > thread 2 > 00007f655a1a6000 2m 4K 512 10 502 > thread stack: > thread 3 > 00007f6559fa5000 2m 4K 512 10 502 > thread stack: > thread 4 > 00007f6559da4000 2m 4K 512 10 502 > thread stack: > thread 5 > 00007f6559ba3000 2m 4K 512 10 502 > thread stack: > thread 6 > 00007f65599a2000 2m 4K 512 10 502 > thread stack: > thread 7 > 00007f65597a1000 2m 4K 512 11 501 > thread stack: > thread 8 > 00007f65595a0000 2m 4K 512 10 502 > thread stack: > thread 9 > 00007f655939f000 2m 4K 512 10 502 > thread stack: > thread 10 > 00007f655919e000 2m 4K 512 10 502 > thread stack: > thread 11 > 00007f6558f9d000 2m 4K 512 10 502 > thread stack: > thread 12 > 00007f6558d9c000 2m 4K 512 10 502 > thread stack: > thread 13 > 00007f6558b9b000 2m 4K 512 10 502 > thread stack: > thread 14 > 00007f655899a000 2m 4K 512 10 502 > thread stack: > thread 15 > 00007f6558799000 2m 4K 512 10 502 > thread stack: > thread 16 > 00007f6558598000 2m 4K 512 10 502 > thread stack: > thread 17 > 00007f6558397000 2m 4K 512 10 502 > thread stack: > thread 18 > 00007f6558196000 2m 4K 512 10 502 > thread stack: > thread 19 > 00007f6553e00000 2m 4K 512 10 502 > thread stack: > thread 20 > 00007f6553bff000 2m 4K 512 10 502 > thread stack: > thread 21 > 00007f65539fe000 2m 4K 512 10 502 > thread stack: > thread 22 > 00007f65537fd000 2m 4K 512 10 502 > thread stack: > thread 23 > 00007f65535fc000 2m 4K 512 10 502 > thread stack: > thread 24 > 00007f65533fb000 2m 4K 512 10 502 > thread stack: > thread 25 > 00007f65531fa000 2m 4K 512 10 502 > thread stack: > thread 26 > 00007f6552ff9000 2m 4K 512 10 502 > thread stack: > thread 27 > 00007f6552f98000 64k 4K 16 2 14 > process stack: > lacp-process > 00007f6552f57000 256k 4K 64 3 61 > process stack: > unix-cli-local:0 > 00007f6552f46000 64k 4K 16 2 14 > process stack: > unix-cli-new-session > > > > > > > > > > > -- > Paweł Staszewski > CEO at ITCare > > Cell: +48 609911040 > pstaszew...@itcare.pl <mailto:pstaszew...@itcare.pl> > > This message (including any attachments) contains confidential information > intended > for a specific individual and purpose, and is protected by law. If you > are not the > intended recipient, you should delete this message. Any disclosure, > copying, or > distribution of this message, or the taking of any action based on it, is > strictly > prohibited. > > -------------------------------------------------------------------------- > ----------- > > Wiadomość ta jest przeznaczona jedynie dla osoby lub podmiotu będącego jej > adresatem i > może zawierać poufne lub uprzywilejowane informacje. Zakazane jest > przeglądanie, > przesyłanie, rozpowszechnianie lub inne wykorzystywanie tych informacji, > jak również > podejmowanie działań na ich podstawie, przez osoby lub podmioty inne niż > zamierzony > adresat. Jeśli otrzymali Państwo tę wiadomość przez pomyłkę, prosimy o > poinformowanie > nadawcy i usunięcie jej z komputera.
-=-=-=-=-=-=-=-=-=-=-=- Links: You receive all messages sent to this group. View/Reply Online (#21883): https://lists.fd.io/g/vpp-dev/message/21883 Mute This Topic: https://lists.fd.io/mt/93619900/21656 Group Owner: vpp-dev+ow...@lists.fd.io Unsubscribe: https://lists.fd.io/g/vpp-dev/leave/1480452/21656/631435203/xyzzy [arch...@mail-archive.com] -=-=-=-=-=-=-=-=-=-=-=-
