Hi, i have a problem on my 10Gb NIC SolarFlare SFN5162F SFP+. systat -if and netstat -I sfxge not showing out traffic. For my Intel gbit interface em0 systat/netstat showin all correctly. /0 /1 /2 /3 /4 /5 /6 /7 /8 /9 /10 Load Average Interface Traffic Peak Total lo0 in 0.000 KB/s 0.000 KB/s 0.090 KB out 0.000 KB/s 0.000 KB/s 0.090 KB em0 in 20.778 KB/s 22.007 KB/s 2.169 MB out 26.362 KB/s 26.362 KB/s 274.287 KB sfxge1 in 7.598 KB/s 21.739 KB/s 10.926 MB out 0.000 KB/s 0.000 KB/s 0.000 KB # netstat -I em0 -i 1 input em0 output packets errs idrops bytes packets errs bytes colls 19 0 0 14962 22 0 29044 0 16 0 0 14845 21 0 28932 0 20 0 0 15478 22 0 28878 0 24 0 0 15394 21 0 28878 0 15 0 0 14702 21 0 28878 0 # netstat -I sfxge1 -i 1 input sfxge1 output packets errs idrops bytes packets errs bytes colls 0 0 0 15127 0 0 0 0 0 0 0 14884 0 0 0 0 0 0 0 15234 0 0 0 0 0 0 0 14948 0 0 0 0 0 0 0 14768 0 0 0 0 10.2-RELEASE FreeBSD 10.2-RELEASE #1: Sat Sep 5 16:56:24 CEST 2015 Best regards
Not on net, but I figured this bug needs more information before it can be appropriately triaged. Please provide... - ... the ifconfig output for em0/sfxge1. - ... the output of `sysctl hw.sfxge`.
Sidenote: based on code inspection, specifying -i when specifying -I <interface> is redundant.
Hi, sfxge1: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> metric 0 mtu 1500 options=6c07ab<RXCSUM,TXCSUM,VLAN_MTU,JUMBO_MTU,VLAN_HWCSUM,TSO4,TSO6,LRO,VLAN_HWTSO,LINKSTATE,RXCSUM_IPV6,TXCSUM_IPV6> ether 00:0f:53:08:48:89 inet 192.168.2.1 netmask 0xffffff00 broadcast 192.168.2.255 nd6 options=29<PERFORMNUD,IFDISABLED,AUTO_LINKLOCAL> media: Ethernet 10Gbase-SR <full-duplex,rxpause,txpause> status: active em0: flags=8c02<BROADCAST,OACTIVE,SIMPLEX,MULTICAST> metric 0 mtu 1500 options=4019b<RXCSUM,TXCSUM,VLAN_MTU,VLAN_HWTAGGING,VLAN_HWCSUM,TSO4,VLAN_HWTSO> ether 74:d4:35:e7:40:af inet 10.10.2.1 netmask 0xffffff00 broadcast 10.10.2.255 nd6 options=29<PERFORMNUD,IFDISABLED,AUTO_LINKLOCAL> media: Ethernet autoselect (1000baseT <full-duplex>) status: active root@rtr:/# sysctl -a | grep hw.sfxge hw.sfxge.tso_fw_assisted: 1 hw.sfxge.tx_dpl_put_max: 1024 hw.sfxge.tx_dpl_get_non_tcp_max: 1024 hw.sfxge.tx_dpl_get_max: 65536 hw.sfxge.lro.loss_packets: 20 hw.sfxge.lro.slow_start_packets: 2000 hw.sfxge.lro.idle_ticks: 0 hw.sfxge.lro.chain_max: 20 hw.sfxge.lro.table_size: 128 hw.sfxge.tx_ring: 1024 hw.sfxge.rx_ring: 1024 root@rtr:/# sysctl -a | grep hw.em hw.em.eee_setting: 1 hw.em.rx_process_limit: 100 hw.em.enable_msix: 1 hw.em.sbp: 0 hw.em.smart_pwr_down: 0 hw.em.txd: 1024 hw.em.rxd: 1024 hw.em.rx_abs_int_delay: 66 hw.em.tx_abs_int_delay: 66 hw.em.rx_int_delay: 0 hw.em.tx_int_delay: 66 root@rtr:/# sysctl -a | grep octets dev.em.2.mac_stats.good_octets_txd: 0 dev.em.2.mac_stats.good_octets_recvd: 0 dev.em.1.mac_stats.good_octets_txd: 0 dev.em.1.mac_stats.good_octets_recvd: 0 dev.em.0.mac_stats.good_octets_txd: 456051 dev.em.0.mac_stats.good_octets_recvd: 1531824 dev.sfxge.1.stats.tx_octets: 1087206 dev.sfxge.1.stats.rx_octets: 23345200
That's my fault. Since statistics support differs in HEAD (if_get_counter) and earlier stable branches (periodic update of counters in ifnet). I've forgotten to merge stable/10-specific patch. I'll care about it. (If it helps you, there is no the issue in driver available on Solarflare support site)
A commit references this bug: Author: arybchik Date: Fri Dec 11 06:41:02 UTC 2015 New revision: 292089 URL: https://svnweb.freebsd.org/changeset/base/292089 Log: sfxge: implement interface statistics shown by netstat PR: 202907 Submitted by: Boris Misenov <Boris.Misenov at oktetlabs.ru> Reviewed by: gnn, philip Sponsored by: Solarflare Communications, Inc. Differential Revision: https://reviews.freebsd.org/D4317 Changes: stable/10/sys/dev/sfxge/sfxge.c stable/10/sys/dev/sfxge/sfxge.h stable/10/sys/dev/sfxge/sfxge_port.c stable/10/sys/dev/sfxge/sfxge_tx.c stable/10/sys/dev/sfxge/sfxge_tx.h stable/10/sys/dev/sfxge/sfxge_version.h
Fixed in 10.3-RELEASE