LINUX.ORG.RU
ФорумAdmin

Проблемы с Jumbo Frames на RTL8111E

 , ,


1

2

Я сегодня случайно обнаружил, что после запуска гипервайзера не проходят большие пакеты. В iperf3 такая хрень:

ne-vlezay80@ne-vlezay80:~$ iperf3 -c 10.243.0.253
Connecting to host 10.243.0.253, port 5201
[  4] local 10.247.1.20 port 42916 connected to 10.243.0.253 port 5201
[ ID] Interval           Transfer     Bandwidth       Retr  Cwnd
[  4]   0.00-1.00   sec   175 KBytes  1.43 Mbits/sec    0   8.74 KBytes       
[  4]   1.00-2.00   sec  0.00 Bytes  0.00 bits/sec    0   8.74 KBytes       
[  4]   2.00-3.00   sec  0.00 Bytes  0.00 bits/sec    0   8.74 KBytes       
[  4]   3.00-4.00   sec  0.00 Bytes  0.00 bits/sec    0   8.74 KBytes       
[  4]   4.00-5.00   sec  0.00 Bytes  0.00 bits/sec    0   8.74 KBytes       
[  4]   5.00-6.00   sec  0.00 Bytes  0.00 bits/sec    0   8.74 KBytes       
[  4]   6.00-7.00   sec  0.00 Bytes  0.00 bits/sec    0   8.74 KBytes       
[  4]   7.00-8.00   sec  0.00 Bytes  0.00 bits/sec    0   8.74 KBytes       
^C[  4]   8.00-8.43   sec  0.00 Bytes  0.00 bits/sec    0   8.74 KBytes       
- - - - - - - - - - - - - - - - - - - - - - - - -
[ ID] Interval           Transfer     Bandwidth       Retr
[  4]   0.00-8.43   sec   175 KBytes   170 Kbits/sec    0             sender
[  4]   0.00-8.43   sec  0.00 Bytes  0.00 bits/sec                  receiver
iperf3: interrupt - the client has terminated
ne-vlezay80@ne-vlezay80:~$ iperf3 -c 10.243.0.253
^C- - - - - - - - - - - - - - - - - - - - - - - - -
[ ID] Interval           Transfer     Bandwidth       Retr
Ошибка сегментирования

Потом в логах ядра:

[  135.609457] iperf3[3151]: segfault at 20 ip 00007f4482a77c7e sp 00007fffebb10df0 error 4 in libiperf.so.0.0.0[7f4482a6a000+1c000]
[  139.048338] iperf3[3154]: segfault at 20 ip 00007f56abefdc7e sp 00007ffd39c1ae40 error 4 in libiperf.so.0.0.0[7f56abef0000+1c000]
[  156.845542] ------------[ cut here ]------------
[  156.845562] WARNING: CPU: 0 PID: 0 at net/sched/sch_generic.c:272 dev_watchdog+0x210/0x220
[  156.845565] NETDEV WATCHDOG: eth0 (r8169): transmit queue 0 timed out
[  156.845572] Modules linked in: vhost_net vhost macvtap macvlan af_packet tun ebt_ip ebt_ip6 ebt_vlan ebtable_filter ebtables x_tables dummy nbd bridge 8021q garp mrp stp llc ata_generic pata_acpi snd_hda_codec_realtek snd_hda_codec_generic snd_usb_audio snd_usbmidi_lib uvcvideo snd_rawmidi snd_seq_device videobuf2_vmalloc videobuf2_memops videobuf2_v4l2 videobuf2_core videodev media kvm_amd kvm irqbypass nouveau snd_hda_intel snd_hda_codec snd_pcsp snd_hda_core snd_hwdep snd_pcm snd_timer snd r8169 soundcore ttm mii wmi pata_atiixp tpm_infineon ohci_pci ohci_hcd acpi_cpufreq tpm_tis tpm fuse ipv6 unix
[  156.845604] CPU: 0 PID: 0 Comm: swapper/0 Not tainted 4.7.1 #1
[  156.845606] Hardware name: MSI MS-7715/870-C45(FX) V2 (MS-7715)  , BIOS V3.1 04/16/2012
[  156.845607]  0000000000000000 ffff88012b003d98 ffffffff81344f08 ffff88012b003de8
[  156.845612]  0000000000000000 ffff88012b003dd8 ffffffff8105a8ec 0000011000000005
[  156.845615]  0000000000000000 ffff8800cb8c2000 0000000000000000 0000000000000001
[  156.845619] Call Trace:
[  156.845620]  <IRQ>  [<ffffffff81344f08>] dump_stack+0x4d/0x65
[  156.845628]  [<ffffffff8105a8ec>] __warn+0xcc/0xf0
[  156.845631]  [<ffffffff8105a95a>] warn_slowpath_fmt+0x4a/0x50
[  156.845634]  [<ffffffff810715ae>] ? __queue_work+0x12e/0x400
[  156.845636]  [<ffffffff81607a30>] dev_watchdog+0x210/0x220
[  156.845638]  [<ffffffff81607820>] ? dev_deactivate_queue.constprop.32+0x60/0x60
[  156.845643]  [<ffffffff810b3c60>] call_timer_fn+0x30/0x130
[  156.845646]  [<ffffffff81607820>] ? dev_deactivate_queue.constprop.32+0x60/0x60
[  156.845648]  [<ffffffff810b4357>] run_timer_softirq+0x1f7/0x2a0
[  156.845650]  [<ffffffff8105fe18>] __do_softirq+0xf8/0x280
[  156.845652]  [<ffffffff810600c3>] irq_exit+0x83/0x90
[  156.845657]  [<ffffffff8101e48f>] do_IRQ+0x4f/0xd0
[  156.845661]  [<ffffffff81759fbf>] common_interrupt+0x7f/0x7f
[  156.845662]  <EOI>  [<ffffffff81025e96>] ? default_idle+0x16/0xd0
[  156.845666]  [<ffffffff81025f84>] amd_e400_idle+0x34/0xf0
[  156.845668]  [<ffffffff8102667a>] arch_cpu_idle+0xa/0x10
[  156.845670]  [<ffffffff81094520>] default_idle_call+0x30/0x40
[  156.845671]  [<ffffffff810947f5>] cpu_startup_entry+0x2c5/0x350
[  156.845673]  [<ffffffff81753902>] rest_init+0x72/0x80
[  156.845676]  [<ffffffff81d3ee78>] start_kernel+0x3e2/0x3ef
[  156.845678]  [<ffffffff81d3e434>] x86_64_start_reservations+0x2f/0x31
[  156.845682]  [<ffffffff81d3e520>] x86_64_start_kernel+0xea/0xed
[  156.845685] ---[ end trace 9600325ec049f29e ]---
[  156.855762] r8169 0000:02:00.0 eth0: link up
Ещё вот ping:
Ping контейнера на виртуалке во время сбоя:
ne-vlezay80@ne-vlezay80:~$ ping 10.243.0.253
PING 10.243.0.253 (10.243.0.253) 56(84) bytes of data.
From 10.247.1.1 icmp_seq=1 Destination Host Unreachable
From 10.247.1.1 icmp_seq=2 Destination Host Unreachable
From 10.247.1.1 icmp_seq=3 Destination Host Unreachable
From 10.247.1.1 icmp_seq=4 Destination Host Unreachable
From 10.247.1.1 icmp_seq=5 Destination Host Unreachable
From 10.247.1.1 icmp_seq=6 Destination Host Unreachable
From 10.247.1.1 icmp_seq=7 Destination Host Unreachable
From 10.247.1.1 icmp_seq=8 Destination Host Unreachable
From 10.247.1.1 icmp_seq=9 Destination Host Unreachable
From 10.247.1.1 icmp_seq=10 Destination Host Unreachable
64 bytes from 10.243.0.253: icmp_seq=11 ttl=61 time=1460 ms
64 bytes from 10.243.0.253: icmp_seq=12 ttl=61 time=461 ms
64 bytes from 10.243.0.253: icmp_seq=13 ttl=61 time=0.486 ms
64 bytes from 10.243.0.253: icmp_seq=14 ttl=61 time=0.313 ms
64 bytes from 10.243.0.253: icmp_seq=15 ttl=61 time=0.436 ms
64 bytes from 10.243.0.253: icmp_seq=16 ttl=61 time=0.427 ms
64 bytes from 10.243.0.253: icmp_seq=17 ttl=61 time=0.390 ms
64 bytes from 10.243.0.253: icmp_seq=18 ttl=61 time=0.369 ms
64 bytes from 10.243.0.253: icmp_seq=19 ttl=61 time=0.365 ms
64 bytes from 10.243.0.253: icmp_seq=20 ttl=61 time=0.351 ms
^C
--- 10.243.0.253 ping statistics ---
20 packets transmitted, 10 received, +10 errors, 50% packet loss, time 19006ms
rtt min/avg/max/mdev = 0.313/192.564/1460.794/444.554 ms, pipe 3
Сам сбой появляется только после загрузки хоста.
ifconfig:
root@ne-vlezay80:/home/ne-vlezay80# ifconfig eth0
eth0      Link encap:Ethernet  HWaddr 8c:89:a5:9d:fd:19  
          inet addr:10.247.1.20  Bcast:10.247.1.255  Mask:255.255.255.0
          inet6 addr: fe80::8e89:a5ff:fe9d:fd19/64 Scope:Link
          UP BROADCAST RUNNING MULTICAST  MTU:9000  Metric:1
          RX packets:625735 errors:0 dropped:0 overruns:0 frame:0
          TX packets:625544 errors:0 dropped:47 overruns:0 carrier:0
          collisions:0 txqueuelen:1000 
          RX bytes:4111173820 (3.8 GiB)  TX bytes:4112784642 (3.8 GiB)

Модуль ядра: r8169

★★★★★

А не факт что он вообще их умеет в размере 9000 байт. Этож Realtek. Попробуйте выставить mtu в 7000 байт.

trancefer ★★ ()
Вы не можете добавлять комментарии в эту тему. Тема перемещена в архив.