| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-ID: <a59c1fdf-9c32-50f6-eb45-040e4ce30cff@itcare.pl>
Date: Sat, 5 Aug 2017 18:38:11 +0200
From: Paweł Staszewski <pstaszewski@...are.pl>
To: netdev@...r.kernel.org
Subject: Performance comparision / qdisc / no qdisc / queueing linux kernel
4.12
Hi
Want to send some comparisions about poor performance of kernel when
simple queueing is enabled
So first - kernel performance for forwarded udp packets generated from
pktgen:
UDP stream ~10Mpps - random destination from network 172.16.0.0/12 +
random udp port
Forwarding performance with 8x RSS queues
enp216s0f0: 9772859.00 P/s 0.00 P/s 9772859.00 P/s
enp216s0f1: 0.00 P/s 9772883.00 P/s 9772883.00 P/s
------------------------------------------------------------------------------
total: 9772863.00 P/s 9772886.00 P/s 19545750.00 P/s
mpstat -P ALL 1 10
Average: CPU %usr %nice %sys %iowait %irq %soft %steal
%guest %gnice %idle
Average: all 0.00 0.00 0.00 0.00 0.00 11.89
0.00 0.00 0.00 88.11
Average: 0 0.00 0.00 0.00 0.00 0.00 0.10 0.00
0.00 0.00 99.90
Average: 1 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 100.00
Average: 2 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 100.00
Average: 3 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 100.00
Average: 4 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 100.00
Average: 5 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 100.00
Average: 6 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 100.00
Average: 7 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 100.00
Average: 8 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 100.00
Average: 9 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 100.00
Average: 10 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 100.00
Average: 11 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 100.00
Average: 12 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 100.00
Average: 13 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 100.00
Average: 14 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 100.00
Average: 15 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 100.00
Average: 16 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 100.00
Average: 17 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 100.00
Average: 18 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 100.00
Average: 19 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 100.00
Average: 20 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 100.00
Average: 21 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 100.00
Average: 22 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 100.00
Average: 23 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 100.00
Average: 24 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 100.00
Average: 25 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 100.00
Average: 26 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 100.00
Average: 27 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 100.00
Average: 28 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 100.00
Average: 29 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 100.00
Average: 30 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 100.00
Average: 31 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 100.00
Average: 32 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 100.00
Average: 33 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 100.00
Average: 34 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 100.00
Average: 35 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 100.00
Average: 36 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 100.00
Average: 37 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 100.00
Average: 38 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 100.00
Average: 39 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 100.00
Average: 40 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 100.00
Average: 41 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 100.00
Average: 42 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 100.00
Average: 43 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 100.00
Average: 44 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 100.00
Average: 45 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 100.00
Average: 46 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 100.00
Average: 47 0.00 0.00 0.00 0.00 0.00 0.00 0.00
0.00 0.00 100.00
Average: 48 0.00 0.00 0.00 0.00 0.00 84.10
0.00 0.00 0.00 15.90
Average: 49 0.00 0.00 0.00 0.00 0.00 83.20
0.00 0.00 0.00 16.80
Average: 50 0.00 0.00 0.00 0.00 0.00 81.98
0.00 0.00 0.00 18.02
Average: 51 0.00 0.00 0.00 0.00 0.00 80.80
0.00 0.00 0.00 19.20
Average: 52 0.00 0.00 0.00 0.00 0.00 82.20
0.00 0.00 0.00 17.80
Average: 53 0.00 0.00 0.00 0.00 0.00 85.30
0.00 0.00 0.00 14.70
Average: 54 0.00 0.00 0.00 0.00 0.00 85.30
0.00 0.00 0.00 14.70
Average: 55 0.00 0.00 0.00 0.00 0.00 82.88
0.00 0.00 0.00 17.12
So there is no problem with 8x queues to forward almost 10Mpps with ~90%
cpu load on 8x queues
some turbostat:
./turbostat
turbostat version 17.06.23 - Len Brown <lenb@...nel.org>
CPUID(0): GenuineIntel 22 CPUID levels; family:model:stepping 0x6:55:4
(6:85:4)
CPUID(1): SSE3 MONITOR SMX EIST TM2 TSC MSR ACPI-TM TM
CPUID(6): APERF, TURBO, DTS, PTM, No-HWP, No-HWPnotify, No-HWPwindow,
No-HWPepp, No-HWPpkg, EPB
cpu12: MSR_IA32_MISC_ENABLE: 0x00850089 (TCC EIST No-MWAIT PREFETCH TURBO)
CPUID(7): No-SGX
CPUID(0x15): eax_crystal: 2 ebx_tsc: 208 ecx_crystal_hz: 0
TSC: 2600 MHz (25000000 Hz * 208 / 2 / 1000000)
CPUID(0x16): base_mhz: 2600 max_mhz: 3700 bus_mhz: 100
cpu12: MSR_MISC_PWR_MGMT: 0x00402000 (ENable-EIST_Coordination
DISable-EPB DISable-OOB)
RAPL: 1872 sec. Joule Counter Range, at 140 Watts
cpu12: MSR_PLATFORM_INFO: 0x70a2cf3811a00
10 * 100.0 = 1000.0 MHz max efficiency frequency
26 * 100.0 = 2600.0 MHz base frequency
cpu12: MSR_IA32_POWER_CTL: 0x2904005b (C1E auto-promotion: ENabled)
cpu12: MSR_TURBO_RATIO_LIMIT: 0x2121212122222325
cpu12: MSR_TURBO_RATIO_LIMIT1: 0x1c18140e0c080402
33 * 100.0 = 3300.0 MHz max turbo 28 active cores
33 * 100.0 = 3300.0 MHz max turbo 24 active cores
33 * 100.0 = 3300.0 MHz max turbo 20 active cores
33 * 100.0 = 3300.0 MHz max turbo 14 active cores
34 * 100.0 = 3400.0 MHz max turbo 12 active cores
34 * 100.0 = 3400.0 MHz max turbo 8 active cores
35 * 100.0 = 3500.0 MHz max turbo 4 active cores
37 * 100.0 = 3700.0 MHz max turbo 2 active cores
cpu12: MSR_CONFIG_TDP_NOMINAL: 0x0000001a (base_ratio=26)
cpu12: MSR_CONFIG_TDP_LEVEL_1: 0x108094800110460 (PKG_MIN_PWR_LVL1=264
PKG_MAX_PWR_LVL1=2376 LVL1_RATIO=17 PKG_TDP_LVL1=1120)
cpu12: MSR_CONFIG_TDP_LEVEL_2: 0x108094800110460 (PKG_MIN_PWR_LVL2=264
PKG_MAX_PWR_LVL2=2376 LVL2_RATIO=17 PKG_TDP_LVL2=1120)
cpu12: MSR_CONFIG_TDP_CONTROL: 0x80000000 ( lock=1)
cpu12: MSR_TURBO_ACTIVATION_RATIO: 0x000000ff (MAX_NON_TURBO_RATIO=255
lock=0)
cpu12: MSR_PKG_CST_CONFIG_CONTROL: 0x00008403 (locked:
pkg-cstate-limit=3: pc6r)
cpu12: POLL: CPUIDLE CORE POLL IDLE
cpu12: C1: ACPI FFH INTEL MWAIT 0x0
cpu12: C2: ACPI FFH INTEL MWAIT 0x20
cpu12: cpufreq driver: acpi-cpufreq
cpu12: cpufreq governor: performance
cpufreq boost: 1
cpu12: MSR_MISC_FEATURE_CONTROL: 0x00000000 (L2-Prefetch
L2-Prefetch-pair L1-Prefetch L1-IP-Prefetch)
cpu0: MSR_IA32_ENERGY_PERF_BIAS: 0x00000006 (balanced)
cpu14: MSR_IA32_ENERGY_PERF_BIAS: 0x00000006 (balanced)
cpu0: MSR_RAPL_POWER_UNIT: 0x000a0e03 (0.125000 Watts, 0.000061 Joules,
0.000977 sec.)
cpu0: MSR_PKG_POWER_INFO: 0xf094802100460 (140 W TDP, RAPL 66 - 297 W,
0.014648 sec.)
cpu0: MSR_PKG_POWER_LIMIT: 0x3854000148460 (UNlocked)
cpu0: PKG Limit #1: ENabled (140.000000 Watts, 1.000000 sec, clamp DISabled)
cpu0: PKG Limit #2: ENabled (168.000000 Watts, 0.001953* sec, clamp ENabled)
cpu0: MSR_DRAM_POWER_INFO,: 0xf00a000120098 (19 W TDP, RAPL 2 - 20 W,
0.014648 sec.)
cpu0: MSR_DRAM_POWER_LIMIT: 0x00000000 (UNlocked)
cpu0: DRAM Limit: DISabled (0.000000 Watts, 0.000977 sec, clamp DISabled)
cpu14: MSR_RAPL_POWER_UNIT: 0x000a0e03 (0.125000 Watts, 0.000061 Joules,
0.000977 sec.)
cpu14: MSR_PKG_POWER_INFO: 0xf094802100460 (140 W TDP, RAPL 66 - 297 W,
0.014648 sec.)
cpu14: MSR_PKG_POWER_LIMIT: 0x3854000148460 (UNlocked)
cpu14: PKG Limit #1: ENabled (140.000000 Watts, 1.000000 sec, clamp
DISabled)
cpu14: PKG Limit #2: ENabled (168.000000 Watts, 0.001953* sec, clamp
ENabled)
cpu14: MSR_DRAM_POWER_INFO,: 0xf00a000120098 (19 W TDP, RAPL 2 - 20 W,
0.014648 sec.)
cpu14: MSR_DRAM_POWER_LIMIT: 0x00000000 (UNlocked)
cpu14: DRAM Limit: DISabled (0.000000 Watts, 0.000977 sec, clamp DISabled)
cpu0: MSR_IA32_TEMPERATURE_TARGET: 0x00650a00 (101 C)
cpu14: MSR_IA32_TEMPERATURE_TARGET: 0x00650a00 (101 C)
cpu0: MSR_IA32_PACKAGE_THERM_STATUS: 0x882f0800 (54 C)
cpu0: MSR_IA32_PACKAGE_THERM_INTERRUPT: 0x00000003 (101 C, 101 C)
cpu14: MSR_IA32_PACKAGE_THERM_STATUS: 0x88250800 (64 C)
cpu14: MSR_IA32_PACKAGE_THERM_INTERRUPT: 0x00000003 (101 C, 101 C)
cpu12: MSR_PKGC3_IRTL: 0x00000000 (NOTvalid, 0 ns)
cpu12: MSR_PKGC6_IRTL: 0x00000000 (NOTvalid, 0 ns)
cpu12: MSR_PKGC7_IRTL: 0x00000000 (NOTvalid, 0 ns)
Package Core CPU Avg_MHz Busy% Bzy_MHz TSC_MHz IRQ SMI
C1 C2 C1% C2% CPU%c1 CPU%c6 CoreTmp PkgTmp Pkg%pc2
Pkg%pc6 PkgWatt RAMWatt PKG_% RAM_%
- - - 433 13.25 3265 2600 3383743 0
111 1620625 0.09 86.94 26.20 60.55 -4 64 42.78
0.00 168.80 20.11 0.00 0.00
0 0 0 8 0.78 1029 2600 1650 0
71 1547 2.25 97.19 12.97 86.25 50 54 85.55
0.00 60.31 5.64 0.00 0.00
0 0 28 7 0.69 1018 2600 1255 0
16 1246 1.28 98.26 13.07
0 1 1 7 0.70 1020 2600 1260 0
0 1257 0.00 99.52 9.29 90.01 53
0 1 29 7 0.66 1018 2600 1255 0
0 1256 0.00 99.52 9.33
0 2 2 7 0.68 1019 2600 1260 0
0 1259 0.00 99.51 9.31 90.01 51
0 2 30 7 0.67 1019 2600 1255 0
0 1256 0.00 99.52 9.32
0 3 3 7 0.68 1019 2600 1260 0
0 1259 0.00 99.52 9.33 89.99 -4
0 3 31 7 0.68 1018 2600 1255 0
0 1255 0.00 99.53 9.33
0 4 4 7 0.69 1020 2600 1260 0
0 1259 0.00 99.50 9.35 89.96 -4
0 4 32 7 0.69 1018 2600 1255 0
0 1255 0.00 99.51 9.35
0 5 5 7 0.69 1021 2600 1260 0
0 1259 0.00 99.52 9.31 89.99 53
0 5 33 7 0.70 1021 2600 1255 0
0 1255 0.00 99.52 9.31
0 6 6 7 0.71 1022 2600 1260 0
0 1259 0.00 99.52 9.31 89.97 -4
0 6 34 7 0.73 1022 2600 1255 0
0 1255 0.00 99.52 9.30
0 8 7 7 0.71 1021 2600 1257 0
4 1253 0.32 99.20 9.52 89.77 54
0 8 35 7 0.67 1018 2600 1255 0
0 1255 0.00 99.52 9.56
0 9 8 7 0.68 1019 2600 1266 0
0 1258 0.00 99.51 9.53 89.79 52
0 9 36 7 0.68 1017 2600 1255 0
4 1250 0.32 99.20 9.53
0 10 9 7 0.67 1020 2600 1257 0
5 1253 0.24 99.29 9.55 89.78 -4
0 10 37 7 0.68 1018 2600 1255 0
4 1250 0.32 99.21 9.55
0 11 10 7 0.69 1018 2600 1257 0
0 1258 0.00 99.51 9.31 89.99 -4
0 11 38 7 0.68 1018 2600 1255 0
0 1254 0.00 99.52 9.33
0 12 11 7 0.69 1020 2600 1257 0
0 1257 0.00 99.52 9.31 90.00 -4
0 12 39 7 0.69 1019 2600 1255 0
0 1254 0.00 99.53 9.31
0 13 12 9 0.77 1182 2600 1252 0
0 1252 0.00 99.45 9.32 89.91 -4
0 13 40 7 0.71 1021 2600 1255 0
0 1256 0.00 99.53 9.38
0 14 13 7 0.73 1021 2600 1257 0
0 1258 0.00 99.51 9.30 89.97 -4
0 14 41 7 0.73 1019 2600 1255 0
0 1254 0.00 99.52 9.30
1 0 14 10 0.30 3301 2600 1264 0
4 1262 0.24 99.56 26.54 73.16 52 64 0.00
0.00 108.49 14.47 0.00 0.00
1 0 42 9 0.28 3301 2600 1255 0
0 1254 0.00 99.80 26.56
1 1 15 9 0.29 3300 2600 1257 0
0 1258 0.00 99.81 26.34 73.37 56
1 1 43 9 0.27 3301 2600 1255 0
0 1254 0.00 99.81 26.36
1 2 16 9 0.28 3301 2600 1257 0
0 1256 0.00 99.81 26.34 73.39 52
1 2 44 9 0.28 3301 2600 1255 0
0 1254 0.00 99.81 26.34
1 3 17 9 0.29 3301 2600 1257 0
0 1256 0.00 99.81 26.36 73.35 50
1 3 45 9 0.27 3301 2600 1255 0
0 1254 0.00 99.81 26.37
1 4 18 8 0.25 3301 2600 1257 0
0 1256 0.00 99.83 26.32 73.43 54
1 4 46 9 0.26 3301 2600 1255 0
0 1254 0.00 99.82 26.31
1 5 19 9 0.26 3301 2600 1257 0
0 1256 0.00 99.83 26.36 73.38 53
1 5 47 9 0.28 3301 2600 1255 0
0 1255 0.00 99.82 26.35
1 6 20 4 0.10 3400 2600 1257 0
3 1254 0.00 99.92 99.90 0.00 64
1 6 48 3038 90.07 3373 2600 398503 0
0 187917 0.00 10.98 9.93
1 8 21 4 0.11 3399 2600 1257 0
0 1257 0.00 99.90 99.89 0.00 64
1 8 49 2999 88.90 3373 2600 432524 0
0 207412 0.00 12.21 11.10
1 9 22 4 0.10 3400 2600 1257 0
0 1257 0.00 99.91 99.90 0.00 61
1 9 50 2972 88.12 3373 2600 458194 0
0 221818 0.00 13.24 11.88
1 10 23 4 0.11 3399 2600 1257 0
0 1257 0.00 99.91 99.89 0.00 63
1 10 51 3008 89.19 3373 2600 426511 0
0 203832 0.00 11.97 10.81
1 11 24 4 0.10 3399 2600 1257 0
0 1257 0.00 99.91 99.90 0.00 62
1 11 52 2989 88.62 3373 2600 496267 0
0 226496 0.00 12.71 11.38
1 12 25 4 0.11 3297 2600 1257 0
0 1257 0.00 99.91 99.89 0.00 62
1 12 53 2930 91.90 3189 2600 348610 0
0 157897 0.00 8.76 8.10
1 13 26 4 0.14 3237 2600 1258 0
0 1260 0.00 99.88 99.86 0.00 59
1 13 54 2923 91.67 3189 2600 356448 0
0 162567 0.00 9.03 8.33
1 14 27 4 0.11 3399 2600 1257 0
0 1257 0.00 99.91 99.89 0.00 62
1 14 55 3033 89.92 3373 2600 405967 0
0 192122 0.00 11.05 10.08
And perf top
PerfTop: 55080 irqs/sec kernel:99.8% exact: 0.0% [4000Hz
cycles], (all, 56 CPUs)
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
21.93% [kernel] [k] do_raw_spin_lock
7.22% [kernel] [k] fib_table_lookup
6.79% [kernel] [k] acpi_processor_ffh_cstate_enter
5.46% [kernel] [k] ixgbe_poll
4.98% [kernel] [k] netif_receive_skb_internal
2.90% [kernel] [k] __build_skb
2.82% [kernel] [k] ipt_do_table
2.44% [kernel] [k] ixgbe_xmit_frame_ring
2.42% [kernel] [k] ip_rcv
2.31% [kernel] [k] get_rps_cpu
1.98% [kernel] [k] ip_finish_output2
1.87% [kernel] [k] __dev_queue_xmit
1.82% [kernel] [k] __netif_receive_skb_core
1.57% [kernel] [k] ip_route_input_noref
1.39% [kernel] [k] kmem_cache_alloc
1.29% [kernel] [k] ip_forward
1.27% [kernel] [k] irq_entries_start
1.27% [kernel] [k] skb_release_data
1.15% [kernel] [k] netif_skb_features
1.14% [kernel] [k] sch_direct_xmit
1.12% [kernel] [k] page_frag_free
0.99% [kernel] [k] __local_bh_enable_ip
0.78% [kernel] [k] dev_gro_receive
0.77% [kernel] [k] fib_validate_source
0.71% [kernel] [k] ip_rcv_finish
0.71% [kernel] [k] get_dma_ops
0.70% [kernel] [k] __netdev_pick_tx
0.68% [kernel] [k] dev_hard_start_xmit
0.66% [kernel] [k] deliver_ptype_list_skb
0.52% [kernel] [k] ixgbe_alloc_rx_buffers
0.51% [kernel] [k] netdev_pick_tx
0.48% [kernel] [k] cpuidle_enter_state
0.48% [kernel] [k] eth_type_trans
0.47% [kernel] [k] nf_hook_slow
0.46% [kernel] [k] build_skb
0.44% [kernel] [k] is_swiotlb_buffer
And now I will enable simple 10Gbit queue on TX device:
RX NIC:
7: enp216s0f0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state
UP mode DEFAULT group default qlen 8192
link/ether 0c:c4:7a:ea:21:44 brd ff:ff:ff:ff:ff:ff
TX NIC
8: enp216s0f1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state
UP mode DEFAULT group default qlen 8192
link/ether 0c:c4:7a:ea:21:45 brd ff:ff:ff:ff:ff:ff
tc qdisc del dev enp216s0f1 root
tc qdisc add dev enp216s0f1 handle 1: root hfsc default 100
tc class add dev enp216s0f1 parent 1: classid 1:100 hfsc ls m2 20000Mbit
ul m2 20000Mbit
bwm-ng v0.6.1 (probing every 0.500s), press 'h' for help
input: /proc/net/dev type: rate
\ iface Rx Tx Total
==============================================================================
enp24s0f3: 58.00 P/s 2.00 P/s
60.00 P/s
lo: 0.00 P/s 0.00 P/s
0.00 P/s
enp216s0f0: 1119744.00 P/s 0.00 P/s 1119744.00 P/s
enp216s0f1: 0.00 P/s 1090710.00 P/s 1090710.00 P/s
------------------------------------------------------------------------------
total: 1119802.00 P/s 1090712.00 P/s 2210514.00 P/s
Performance drops by almost 90%
from almost 10Mpps to only 1Mpps
perf top
PerfTop: 51808 irqs/sec kernel:99.7% exact: 0.0% [4000Hz
cycles], (all, 56 CPUs)
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
69.02% [kernel] [k] queued_spin_lock_slowpath
5.41% [kernel] [k] do_raw_spin_lock
2.06% [kernel] [k] acpi_processor_ffh_cstate_enter
1.73% [kernel] [k] qdisc_dequeue_head
1.66% [kernel] [k] ip_finish_output2
1.53% [kernel] [k] hfsc_enqueue
1.21% [kernel] [k] __dev_queue_xmit
1.05% [kernel] [k] ixgbe_poll
1.03% [kernel] [k] netif_skb_features
0.84% [kernel] [k] hfsc_dequeue
0.79% [kernel] [k] fib_table_lookup
0.76% [kernel] [k] hfsc_find_class
0.70% [kernel] [k] pfifo_enqueue
0.63% [kernel] [k] sch_direct_xmit
0.53% [kernel] [k] rb_first
0.51% [kernel] [k] skb_release_data
0.48% [kernel] [k] __page_frag_cache_drain
0.44% [kernel] [k] ixgbe_xmit_frame_ring
0.43% [kernel] [k] napi_consume_skb
0.42% [kernel] [k] bstats_update
0.37% [kernel] [k] __qdisc_run
0.33% [kernel] [k] ipt_do_table
0.32% [kernel] [k] __build_skb
0.31% [kernel] [k] __rmqueue_smallest
0.29% [kernel] [k] __free_one_page
0.25% [kernel] [k] ip_rcv
0.25% [kernel] [k] page_frag_free
0.23% [kernel] [k] get_page_from_freelist
0.21% [kernel] [k] __netif_receive_skb_core
0.19% [kernel] [k] read_tsc
0.17% [kernel] [k] kmem_cache_alloc
0.15% [kernel] [k] ip_route_input_noref
0.14% [kernel] [k] get_rps_cpu
0.14% [kernel] [k] ip_forward
0.13% [kernel] [k] update_vf.isra.10
0.10% [kernel] [k] __slab_free
Other settings:
ethtool -x enp216s0f1
RX flow hash indirection table for enp216s0f1 with 8 RX ring(s):
0: 0 1 2 3 4 5 6 7
8: 0 1 2 3 4 5 6 7
16: 0 1 2 3 4 5 6 7
24: 0 1 2 3 4 5 6 7
32: 0 1 2 3 4 5 6 7
40: 0 1 2 3 4 5 6 7
48: 0 1 2 3 4 5 6 7
56: 0 1 2 3 4 5 6 7
64: 0 1 2 3 4 5 6 7
72: 0 1 2 3 4 5 6 7
80: 0 1 2 3 4 5 6 7
88: 0 1 2 3 4 5 6 7
96: 0 1 2 3 4 5 6 7
104: 0 1 2 3 4 5 6 7
112: 0 1 2 3 4 5 6 7
120: 0 1 2 3 4 5 6 7
RSS hash key:
a6:f7:7c:d5:65:99:96:e3:8e:6c:eb:7c:68:1a:3b:02:40:b4:18:27:25:82:25:ba:fb:ba:ee:71:26:a3:f0:20:e4:db:6c:db:af:7e:c5:ee
RSS hash function:
toeplitz: on
xor: off
crc32: off
ethtool -x enp216s0f0
RX flow hash indirection table for enp216s0f0 with 8 RX ring(s):
0: 0 1 2 3 4 5 6 7
8: 0 1 2 3 4 5 6 7
16: 0 1 2 3 4 5 6 7
24: 0 1 2 3 4 5 6 7
32: 0 1 2 3 4 5 6 7
40: 0 1 2 3 4 5 6 7
48: 0 1 2 3 4 5 6 7
56: 0 1 2 3 4 5 6 7
64: 0 1 2 3 4 5 6 7
72: 0 1 2 3 4 5 6 7
80: 0 1 2 3 4 5 6 7
88: 0 1 2 3 4 5 6 7
96: 0 1 2 3 4 5 6 7
104: 0 1 2 3 4 5 6 7
112: 0 1 2 3 4 5 6 7
120: 0 1 2 3 4 5 6 7
RSS hash key:
a6:f7:7c:d5:65:99:96:e3:8e:6c:eb:7c:68:1a:3b:02:40:b4:18:27:25:82:25:ba:fb:ba:ee:71:26:a3:f0:20:e4:db:6c:db:af:7e:c5:ee
RSS hash function:
toeplitz: on
xor: off
crc32: off
for both NIC;s settings:
ifc='enp216s0f0 enp216s0f1'
for i in $ifc
do
ip link set up dev $i
ethtool -A $i autoneg off rx off tx off
ethtool -G $i rx 1024 tx 1024
ip link set $i txqueuelen 8192
ethtool -C $i rx-usecs 10
ethtool -L $i combined 8
ethtool -N $i rx-flow-hash udp4 sdfn
ethtool -K $i ntuple on
done
Some affinity + RPS
./set_irq_affinity.sh -x 48-55 enp216s0f0
./set_irq_affinity.sh -x 48-55 enp216s0f1
So the counters are really bad if queueing is enabled :)
Thanks
Paweł Staszewski
Powered by blists - more mailing lists