Also, I noticed an interesting thing. After I stop the (exit) VPP the queued packets are sent out by the DPDK driver.
On Thu, Aug 19, 2021 at 11:32 AM satish amara <satishkam...@gmail.com> wrote: > What is surprising is everything works fine until I do IPSEC. The ping to > local networks on the eth3 interface works fine until IPSEC traffic starts. > After IPSEC traffic starts on VPP2 the connectivity on eth3 to local LAN > is gone. > > On Thu, Aug 19, 2021 at 11:20 AM Benoit Ganne (bganne) <bga...@cisco.com> > wrote: > >> I do not see any issue in these logs, everything looks like it is working >> fine. >> This could be a config issue on the host side. >> If you use dpdk testpmd between the interfaces instead of VPP, does it >> work? >> Also, you can try using VPP native virtio driver instead of dpdk, see >> "create interface virtio" cli commands. >> >> Best >> ben >> >> > -----Original Message----- >> > From: satish amara <satishkam...@gmail.com> >> > Sent: jeudi 19 août 2021 16:39 >> > To: Benoit Ganne (bganne) <bga...@cisco.com> >> > Cc: vpp-dev@lists.fd.io; Damjan Marion <dmar...@me.com> >> > Subject: Re: [vpp-dev] DPDK initialization issues >> > >> > Hi, >> > Here are the attached logs. I also see the counters on eth3 get >> > increments after pings packets are sent out. The packets are never sent >> > out after that. >> > They are stuck on tx queue. Thank you for looking into this. >> > >> > Regards, >> > Satish K Amara >> > >> > >> > >> > On Thu, Aug 19, 2021 at 3:13 AM Benoit Ganne (bganne) <bga...@cisco.com >> > <mailto:bga...@cisco.com> > wrote: >> > >> > >> > I do not think those are hard errors, especially if packets go >> > through. >> > To debug the ipsec issue we'll need a packet trace, eg.: >> > ~# vppctl cle tr >> > ~# vppctl tr add dpdk-input 100 >> > <reproduce the issue> >> > ~# vppctl sh tr >> > >> > Best >> > ben >> > >> > > -----Original Message----- >> > > From: vpp-dev@lists.fd.io <mailto:vpp-dev@lists.fd.io> <vpp- >> > d...@lists.fd.io <mailto:vpp-dev@lists.fd.io> > On Behalf Of satish >> amara >> > > Sent: mercredi 18 août 2021 20:37 >> > > To: vpp-dev@lists.fd.io <mailto:vpp-dev@lists.fd.io> >> > > Cc: Damjan Marion <dmar...@me.com <mailto:dmar...@me.com> > >> > > Subject: [vpp-dev] DPDK initialization issues >> > > >> > > [Edited Message Follows] >> > > >> > > Hi, >> > > >> > > I tried to bring the VPP with DPDK enabled interface and >> seeing >> > the >> > > following issues. Not sure how to resolve them. >> > > >> > > 2021/08/18 14:12:17:921 notice vat-plug/load Loaded plugin: >> > > nsim_test_plugin.so >> > > 2021/08/18 14:12:17:921 notice dpdk EAL: Detected >> 4 >> > lcore(s) >> > > 2021/08/18 14:12:17:921 notice dpdk EAL: Detected >> 1 >> > NUMA >> > > nodes >> > > 2021/08/18 14:12:17:921 notice dpdk EAL: Selected >> > IOVA mode >> > > 'PA' >> > > 2021/08/18 14:12:17:921 notice dpdk EAL: No >> > available >> > > hugepages reported in hugepages-1048576kB >> > > 2021/08/18 14:12:17:921 notice dpdk EAL: No free >> > hugepages >> > > reported in hugepages-1048576kB >> > > 2021/08/18 14:12:17:921 notice dpdk EAL: No >> > available >> > > hugepages reported in hugepages-1048576kB >> > > 2021/08/18 14:12:17:921 notice dpdk EAL: Probing >> > VFIO >> > > support... >> > > 2021/08/18 14:12:17:921 notice dpdk EAL: VFIO >> > support >> > > initialized >> > > 2021/08/18 14:12:17:921 notice dpdk EAL: WARNING! >> > Base >> > > virtual address hint (0xa80001000 != 0x7fedc0000000) not >> > respected! >> > > 2021/08/18 14:12:17:921 notice dpdk EAL: This >> may >> > cause >> > > issues with mapping memory into secondary processes >> > > 2021/08/18 14:12:17:921 notice dpdk EAL: WARNING! >> > Base >> > > virtual address hint (0xc00002000 != 0x7fe580000000) not >> > respected! >> > > 2021/08/18 14:12:17:921 notice dpdk EAL: This >> may >> > cause >> > > issues with mapping memory into secondary processes >> > > 2021/08/18 14:12:17:921 notice dpdk EAL: WARNING! >> > Base >> > > virtual address hint (0xd80003000 != 0x7fdd40000000) not >> > respected! >> > > 2021/08/18 14:12:17:921 notice dpdk EAL: This >> may >> > cause >> > > issues with mapping memory into secondary processes >> > > 2021/08/18 14:12:17:921 notice dpdk EAL: WARNING! >> > Base >> > > virtual address hi >> > > nt (0xdc0a64000 != 0x7fd93fe00000) not respected! >> > > 2021/08/18 14:12:17:921 notice dpdk EAL: This >> may >> > cause >> > > issues with mapping memory into secondary processes >> > > 2021/08/18 14:12:17:921 notice dpdk EAL: WARNING! >> > Base >> > > virtual address hint (0xdc16c5000 != 0x7fd53fc00000) not >> > respected! >> > > 2021/08/18 14:12:17:921 notice dpdk EAL: This >> may >> > cause >> > > issues with mapping memory into secondary processes >> > > 2021/08/18 14:12:17:921 notice dpdk EAL: WARNING! >> > Base >> > > virtual address hint (0xdc2326000 != 0x7fd13fa00000) not >> > respected! >> > > 2021/08/18 14:12:17:921 notice dpdk EAL: This >> may >> > cause >> > > issues with mapping memory into secondary processes >> > > 2021/08/18 14:12:17:921 notice dpdk EAL: WARNING! >> > Base >> > > virtual address hint (0xdc2f87000 != 0x7fcd3f800000) not >> > respected! >> > > 2021/08/18 14:12:17:921 notice dpdk EAL: This >> may >> > cause >> > > issues with mapping memory into secondary processes >> > > 2021/08/18 14:12:17:921 notice dpdk EAL: Using >> memfd >> > is not >> > > supported, falling back to anonymous hugepages >> > > 2021/08/18 14:12:17:921 notice dpdk EAL: Invalid >> > NUMA >> > > socket, default to 0 >> > > 2021/08/18 14:12:17:921 notice dpdk EAL: Probe PCI >> > driver: >> > > net_virtio (1af4:1000) device: 0000:00:04.0 (socket 0) >> > > 2021/08/18 14:12:17:921 notice dpdk EAL: Invalid >> > NUMA >> > > socket, default to 0 >> > > 2021/08/18 14:12:17:921 notice dpdk EAL: Probe PCI >> > driver: >> > > net_virtio (1af4:1000) device: 0000:00:05.0 (socket 0) >> > > 2021/08/18 14:12:17:921 notice dpdk EAL: Invalid >> > NUMA >> > > socket, default to 0 >> > > 2021/08/18 14:12:17:921 notice dpdk EAL: Probe PCI >> > driver: >> > > net_virtio (1af4:1000) device: 0000:00:06.0 (socket 0) >> > > 2021/08/18 14:12:17:921 notice dpdk EAL: No legacy >> > > callbacks, legacy socket not created >> > > 2021/08/18 14:12:17:921 notice dpdk EAL: VFIO >> > support not >> > > initialized >> > > >> > > cat /etc/sysctl.d/80-vpp.conf >> > > # Number of 2MB hugepages desired >> > > vm.nr_hugepages=1024 >> > > >> > > # Must be greater than or equal to (2 * vm.nr_hugepages). >> > > vm.max_map_count=3096 >> > > >> > > # All groups allowed to access hugepages >> > > vm.hugetlb_shm_group=0 >> > > # Shared Memory Max must be greater or equal to the total size >> of >> > > hugepages. >> > > # For 2MB pages, TotalHugepageSize = vm.nr_hugepages * 2 * 1024 >> * >> > 1024 >> > > # If the existing kernel.shmmax setting (cat >> > /proc/sys/kernel/shmmax) >> > > # is greater than the calculated TotalHugepageSize then set this >> > parameter >> > > # to current shmmax value. >> > > kernel.shmmax=2147483648 >> > > >> > > >> > > >> > > >> > > >> > > But in spite of the error, I was able to ping thru the interface >> > owned by >> > > DPDK. Seeing issue for IPSEC traffic. So trying to find where >> the >> > issue >> > > is >> > > >> > > Thanks, >> > > Satish Amara >> > >> >>
-=-=-=-=-=-=-=-=-=-=-=- Links: You receive all messages sent to this group. View/Reply Online (#19991): https://lists.fd.io/g/vpp-dev/message/19991 Mute This Topic: https://lists.fd.io/mt/84979398/21656 Group Owner: vpp-dev+ow...@lists.fd.io Unsubscribe: https://lists.fd.io/g/vpp-dev/unsub [arch...@mail-archive.com] -=-=-=-=-=-=-=-=-=-=-=-
