Tools for troubleshooting the network performance

X

xiaofan

High Supremacy Member
HWZ Insider
Joined
Sep 16, 2018
Messages
32,622
Reaction score
10,156
bert64 said:
Graphing DNS resolution has a lot of factors which could influence the outcome... Not only are you testing the latency of your connection to the DNS resolver, but also the latency of the connection from the DNS resolver to the authoritative DNS server of the hostname you are querying, plus the authoritative DNS servers in the entire hierarchy from the root all the way up to the actual hostname.
Then you have the added randomness of wether the host you query and/or any of the hierarchy queries are already present in the cache or not.
Click to expand...

I was running smokeping for a while but later decided not to run it. Right now my PVE machine has only one instance of pfSense running, and another LXC container for simple Linux networking utilities. My Asus RT-AX82U router has also the facitlity (EntWare) to run many Linux network utilities.

Just wondering what will be a simple command to be similar to the smokeping dns probe. I was thinking about nslookup or dig. But there seems to be a difference.
 
seowbin

seowbin

Great Supremacy Member
Joined
Oct 10, 2001
Messages
72,913
Reaction score
1,574
what IP should be added to the probes :s13:


just setup smokeping at home
 
X

xiaofan

High Supremacy Member
HWZ Insider
Joined
Sep 16, 2018
Messages
32,622
Reaction score
10,156
Equipment to be used for OOkla SpeedTest -- wired test for internet speed (WAN Speed)

1) Desktop/laptop with 1G/2/5G/10G network adapter -- depending on the plan you are using.

1G/2.5G NICs are now pretty cheap for both desktop and laptops.

10G NICs can be expensive, especially for laptops (Thunderbolt 3 to 10G adapter, or TB3/TB4 dock with 10G NIC).

2) Asus router or some other routers may bundle SpeedTest utility within the router, unfortunately they will not be good enough for 10G.

3) Asus router built-in OOkla SpeedTest utility.
BCM4908 based like Asus RT-AX86U -- not even good for 1Gbps internet (download speed may not be consistent)
BCM4912 based like Asus RT-AX86U Pro or GT-AX6000 -- may be good for 2.5Gbps internet (need confirmation)
BCM4916 based like Asus ROG GT-BE98 or RT-BE96U -- good for 2.5Gbps internet but not 10G internet


RT-AX86U Example results from SIMBA 2.5Gbps thread.
https://forums.hardwarezone.com.sg/threads/simba-fibre-broadband.6883466/page-58#post-151677292

4ntFFmi.png


My RT-AX86U results using Singtel 1Gbps
Qo4SD62.jpg
 
X

xiaofan

High Supremacy Member
HWZ Insider
Joined
Sep 16, 2018
Messages
32,622
Reaction score
10,156
Personally I tend to like OOkla SpeedTest CLI version as it provides a bit more info than the desktop app.

Example wireless test results, Singtel 1Gbps, Asus RT-AX86U, Acer laptop.

Code: 
PS C:\work\speedtest\ookla-speedtest-1.2.0-win64> .\speedtest.exe

   Speedtest by Ookla

      Server: Singtel - Singapore (id: 13623)
         ISP: Singtel Fibre
Idle Latency:     2.53 ms   (jitter: 0.07ms, low: 2.44ms, high: 2.60ms)
    Download:   763.81 Mbps (data used: 631.2 MB)
                 18.02 ms   (jitter: 6.49ms, low: 4.26ms, high: 47.92ms)
      Upload:   656.35 Mbps (data used: 724.0 MB)
                 30.42 ms   (jitter: 12.46ms, low: 7.13ms, high: 197.81ms)
 Packet Loss:     0.0%
  Result URL: https://www.speedtest.net/result/c/89f2e1e2-875c-4ec7-a332-efffa50f4f8f

Same laptop, but with ZTE BE7200 Pro+ wireless AP behind OpenWRT virtual router.

Code: 
PS C:\work\speedtest\ookla-speedtest-1.2.0-win64> .\speedtest.exe

   Speedtest by Ookla

      Server: Singtel - Singapore (id: 13623)
         ISP: Singtel Fibre
Idle Latency:     3.84 ms   (jitter: 0.20ms, low: 3.51ms, high: 3.89ms)
    Download:   903.68 Mbps (data used: 977.9 MB)
                  8.81 ms   (jitter: 3.42ms, low: 3.83ms, high: 30.64ms)
      Upload:   828.15 Mbps (data used: 757.5 MB)
                 16.71 ms   (jitter: 6.16ms, low: 3.25ms, high: 55.30ms)
 Packet Loss:     0.0%
  Result URL: https://www.speedtest.net/result/c/59bd66f2-9650-4fcb-9424-1a40407eb001

Directly from the OpenWRT virtual router.
Code: 
root@OpenWrt:~/speedtest# ./speedtest

   Speedtest by Ookla

      Server: Singtel - Singapore (id: 13623)
         ISP: Singtel Fibre
Idle Latency:     1.47 ms   (jitter: 0.21ms, low: 1.30ms, high: 1.69ms)
    Download:   942.91 Mbps (data used: 422.9 MB)
                  2.80 ms   (jitter: 0.20ms, low: 1.35ms, high: 3.07ms)
      Upload:   946.88 Mbps (data used: 1.1 GB)
                  2.71 ms   (jitter: 0.65ms, low: 1.61ms, high: 16.31ms)
 Packet Loss:     0.0%
  Result URL: https://www.speedtest.net/result/c/7806cc2b-7407-445a-b635-c0eb60757818
 
Last edited:
B

buttertrade69

Senior Member
Joined
Jul 15, 2020
Messages
545
Reaction score
237
xiaofan said:
Personally I tend to like OOkla SpeedTest CLI version as it provides a bit more info than the desktop app.

Example wireless test results, Singtel 1Gbps, Asus RT-AX86U, Acer laptop.

Code: 
PS C:\work\speedtest\ookla-speedtest-1.2.0-win64> .\speedtest.exe

   Speedtest by Ookla

      Server: Singtel - Singapore (id: 13623)
         ISP: Singtel Fibre
Idle Latency:     2.53 ms   (jitter: 0.07ms, low: 2.44ms, high: 2.60ms)
    Download:   763.81 Mbps (data used: 631.2 MB)
                 18.02 ms   (jitter: 6.49ms, low: 4.26ms, high: 47.92ms)
      Upload:   656.35 Mbps (data used: 724.0 MB)
                 30.42 ms   (jitter: 12.46ms, low: 7.13ms, high: 197.81ms)
 Packet Loss:     0.0%
  Result URL: https://www.speedtest.net/result/c/89f2e1e2-875c-4ec7-a332-efffa50f4f8f

Same laptop, but with ZTE BE7200 Pro+ wireless AP behind OpenWRT virtual router.

Code: 
PS C:\work\speedtest\ookla-speedtest-1.2.0-win64> .\speedtest.exe

   Speedtest by Ookla

      Server: Singtel - Singapore (id: 13623)
         ISP: Singtel Fibre
Idle Latency:     3.84 ms   (jitter: 0.20ms, low: 3.51ms, high: 3.89ms)
    Download:   903.68 Mbps (data used: 977.9 MB)
                  8.81 ms   (jitter: 3.42ms, low: 3.83ms, high: 30.64ms)
      Upload:   828.15 Mbps (data used: 757.5 MB)
                 16.71 ms   (jitter: 6.16ms, low: 3.25ms, high: 55.30ms)
 Packet Loss:     0.0%
  Result URL: https://www.speedtest.net/result/c/59bd66f2-9650-4fcb-9424-1a40407eb001

Directly from the OpenWRT virtual router.
Code: 
root@OpenWrt:~/speedtest# ./speedtest

   Speedtest by Ookla

      Server: Singtel - Singapore (id: 13623)
         ISP: Singtel Fibre
Idle Latency:     1.47 ms   (jitter: 0.21ms, low: 1.30ms, high: 1.69ms)
    Download:   942.91 Mbps (data used: 422.9 MB)
                  2.80 ms   (jitter: 0.20ms, low: 1.35ms, high: 3.07ms)
      Upload:   946.88 Mbps (data used: 1.1 GB)
                  2.71 ms   (jitter: 0.65ms, low: 1.61ms, high: 16.31ms)
 Packet Loss:     0.0%
  Result URL: https://www.speedtest.net/result/c/7806cc2b-7407-445a-b635-c0eb60757818
Click to expand...
one thing I don't like about the CLI is you cannot choose the location for the test server. Sometimes want to test how is the speed between my isp to other country I can only do it in GUI.
 
X

xiaofan

High Supremacy Member
HWZ Insider
Joined
Sep 16, 2018
Messages
32,622
Reaction score
10,156
buttertrade69 said:
one thing I don't like about the CLI is you cannot choose the location for the test server. Sometimes want to test how is the speed between my isp to other country I can only do it in GUI.
Click to expand...

You can do that actually. The problem is that you need to know the server ID. There are quite some lists in the Internet but they may keep changing. For example, most of the China servers are dead now.

Eg: https://linuxspeedtest.com/

Code: 
PS C:\work\speedtest\ookla-speedtest-1.2.0-win64> .\speedtest.exe -L
Closest servers:

    ID  Name                           Location             Country
==============================================================================
  4235  StarHub Ltd                    Singapore            Singapore
   367  NewMedia Express               Singapore            Singapore
  7311  M1 Limited                     Singapore            Singapore
  7556  PT FirstMedia                  Singapore            Singapore
 50344  GSL Networks                   Singapore            Singapore
 13623  Singtel                        Singapore            Singapore
  5935  MyRepublic Singapore           Singapore            Singapore
  2054  ViewQwest                      Singapore            Singapore
 45345  Melbikomas UAB                 Singapore            Singapore
 42688  PT XL Axiata Tbk               Singapore            Singapore
PS C:\work\speedtest\ookla-speedtest-1.2.0-win64> .\speedtest.exe -s 2054

   Speedtest by Ookla

      Server: ViewQwest - Singapore (id: 2054)
         ISP: Singtel Fibre
Idle Latency:     4.65 ms   (jitter: 0.47ms, low: 4.02ms, high: 4.87ms)
    Download:   557.48 Mbps (data used: 408.6 MB)
                  6.07 ms   (jitter: 1.81ms, low: 4.15ms, high: 34.81ms)
      Upload:   610.79 Mbps (data used: 1.1 GB)
                 24.46 ms   (jitter: 10.85ms, low: 4.70ms, high: 106.62ms)
 Packet Loss:     0.0%
  Result URL: https://www.speedtest.net/result/c/380c9ca3-064e-4286-9b4c-dc089010bf97
PS C:\work\speedtest\ookla-speedtest-1.2.0-win64> .\speedtest.exe -s 4956

   Speedtest by Ookla

      Server: Celcom Axiata - Kuala Lumpur (id: 4956)
         ISP: Singtel Fibre
Idle Latency:   242.19 ms   (jitter: 30.45ms, low: 232.75ms, high: 279.07ms)
    Download:   398.17 Mbps (data used: 590.5 MB)
                340.00 ms   (jitter: 83.76ms, low: 222.83ms, high: 807.05ms)
      Upload:   228.57 Mbps (data used: 380.9 MB)
                377.79 ms   (jitter: 84.15ms, low: 226.69ms, high: 798.13ms)
 Packet Loss:     0.0%
  Result URL: https://www.speedtest.net/result/c/42617026-647c-4cf6-bd64-c6f76f53671c

Another Reference in this forum.
https://forums.hardwarezone.com.sg/threads/real-international-speed-tests.6405972/
 
X

xiaofan

High Supremacy Member
HWZ Insider
Joined
Sep 16, 2018
Messages
32,622
Reaction score
10,156
xiaofan said:
2) Asus router or some other routers may bundle SpeedTest utility within the router, unfortunately they will not be good enough for 10G.
Click to expand...

From the following post in the HB810 thread. Singtel 10Gbps plan issued TP-Link HB810 is actually pretty good in terms of the built-in OOkla SpeedTest utility.

HB810 is likely to use the same IPQ9574 Quad-Core ARM Cortex A73 CPU as the Archer BE800 and Deco BE85.
https://www.qualcomm.com/products/i...networking-device-finder?Manufacturer=1725358

Reference post in the HB810 thread.
https://forums.hardwarezone.com.sg/...0-wifi-7-router-users.7003498/#post-151425327

S56OMkN.jpeg
 
X

xiaofan

High Supremacy Member
HWZ Insider
Joined
Sep 16, 2018
Messages
32,622
Reaction score
10,156
Alternative to ping: irtt
https://github.com/heistp/irtt

Bash: 
root@ubuntu2204ct11:~# irtt client flent-fremont.bufferbloat.net
[Connecting] connecting to flent-fremont.bufferbloat.net
[23.239.20.41:2112] [Connected] connection established
seq=0 rtt=249.2ms rd=150.8ms sd=98.32ms ipdv=n/a
seq=1 rtt=242.1ms rd=143.6ms sd=98.5ms ipdv=7.02ms
seq=2 rtt=242.8ms rd=144.4ms sd=98.34ms ipdv=637µs
seq=3 rtt=249.1ms rd=150.8ms sd=98.33ms ipdv=6.35ms
seq=4 rtt=252.2ms rd=153.6ms sd=98.56ms ipdv=3.02ms
seq=5 rtt=253.9ms rd=155ms sd=98.82ms ipdv=1.72ms
seq=6 rtt=251.6ms rd=153.3ms sd=98.3ms ipdv=2.25ms
seq=7 rtt=250ms rd=151.7ms sd=98.28ms ipdv=1.64ms
seq=8 rtt=242ms rd=143.5ms sd=98.57ms ipdv=7.96ms
seq=9 rtt=242.7ms rd=143.5ms sd=99.15ms ipdv=646µs
seq=10 rtt=242.6ms rd=144.1ms sd=98.45ms ipdv=75µs
seq=11 rtt=251.7ms rd=153.4ms sd=98.31ms ipdv=9.08ms
seq=12 rtt=256.4ms rd=158.2ms sd=98.26ms ipdv=4.74ms
seq=13 rtt=252.4ms rd=153.4ms sd=99ms ipdv=3.96ms
seq=14 rtt=251.5ms rd=153.4ms sd=98.14ms ipdv=930µs
seq=15 rtt=247.7ms rd=149.3ms sd=98.37ms ipdv=3.78ms
seq=16 rtt=241.8ms rd=143.5ms sd=98.33ms ipdv=5.92ms
seq=17 rtt=245.5ms rd=146.9ms sd=98.63ms ipdv=3.72ms
seq=18 rtt=244.2ms rd=144.6ms sd=99.61ms ipdv=1.32ms
seq=19 rtt=251.8ms rd=153.5ms sd=98.32ms ipdv=7.61ms
seq=20 rtt=252.2ms rd=153.6ms sd=98.61ms ipdv=395µs
seq=21 rtt=251.5ms rd=153.3ms sd=98.25ms ipdv=700µs
seq=22 rtt=251.7ms rd=153.4ms sd=98.25ms ipdv=164µs
seq=23 rtt=246.6ms rd=148.2ms sd=98.42ms ipdv=5.07ms
seq=24 rtt=248.6ms rd=150.1ms sd=98.45ms ipdv=1.95ms
seq=25 rtt=242.4ms rd=143.7ms sd=98.75ms ipdv=6.15ms
seq=26 rtt=245.2ms rd=147ms sd=98.21ms ipdv=2.82ms
seq=28 rtt=252ms rd=153.5ms sd=98.5ms ipdv=n/a
seq=29 rtt=251.7ms rd=153.2ms sd=98.47ms ipdv=320µs
seq=31 rtt=251.9ms rd=153.5ms sd=98.4ms ipdv=n/a
seq=32 rtt=255.4ms rd=153.5ms sd=101.9ms ipdv=3.54ms
seq=33 rtt=249.2ms rd=151ms sd=98.23ms ipdv=6.24ms
seq=34 rtt=242ms rd=143.6ms sd=98.43ms ipdv=7.21ms
seq=35 rtt=242.5ms rd=143.8ms sd=98.72ms ipdv=549µs
seq=36 rtt=250.1ms rd=150.5ms sd=99.67ms ipdv=7.6ms
seq=37 rtt=253ms rd=153.5ms sd=99.49ms ipdv=2.89ms
seq=38 rtt=251.7ms rd=153.4ms sd=98.32ms ipdv=1.36ms
seq=39 rtt=251.9ms rd=153.6ms sd=98.24ms ipdv=209µs
seq=40 rtt=250.5ms rd=152.4ms sd=98.2ms ipdv=1.34ms
seq=41 rtt=242.8ms rd=144.5ms sd=98.28ms ipdv=7.72ms
seq=42 rtt=244.2ms rd=145.7ms sd=98.47ms ipdv=1.33ms
seq=43 rtt=242.3ms rd=144.1ms sd=98.24ms ipdv=1.81ms
seq=44 rtt=243.7ms rd=145.1ms sd=98.64ms ipdv=1.38ms
seq=45 rtt=246.1ms rd=146.4ms sd=99.62ms ipdv=2.34ms
seq=46 rtt=251.8ms rd=153.5ms sd=98.29ms ipdv=5.77ms
seq=47 rtt=251.9ms rd=153.4ms sd=98.58ms ipdv=117µs
seq=48 rtt=251.8ms rd=153.4ms sd=98.38ms ipdv=143µs
seq=49 rtt=251ms rd=152.5ms sd=98.45ms ipdv=806µs
seq=50 rtt=242ms rd=143.5ms sd=98.48ms ipdv=9.02ms
seq=51 rtt=241.9ms rd=143.6ms sd=98.29ms ipdv=89.3µs
seq=52 rtt=242.2ms rd=143.6ms sd=98.6ms ipdv=343µs
seq=53 rtt=251.6ms rd=153.4ms sd=98.19ms ipdv=9.33ms
seq=54 rtt=252.2ms rd=153.7ms sd=98.55ms ipdv=665µs
seq=55 rtt=251.6ms rd=153.4ms sd=98.19ms ipdv=607µs
^Cinterrupt

                         Min     Mean   Median      Max  Stddev
                         ---     ----   ------      ---  ------
                RTT  241.8ms  248.3ms  250.1ms  256.4ms   4.4ms
         send delay  98.14ms  98.58ms  98.43ms  101.9ms   594µs
      receive delay  143.5ms  149.7ms  151.3ms  158.2ms  4.38ms

      IPDV (jitter)     75µs   3.18ms   1.95ms   9.33ms  2.92ms
          send IPDV   5.88µs    489µs    284µs   3.71ms   725µs
       receive IPDV   5.14µs   3.03ms   1.61ms   9.74ms  3.11ms

     send call time   21.8µs   42.2µs            68.2µs  7.47µs
        timer error    2.9µs     79µs             250µs  54.5µs
  server proc. time   10.3µs     48µs             313µs  66.3µs

                duration: 56.4s (wait 0s)
   packets sent/received: 56/54 (3.57% loss)
 server packets received: 56/56 (0.00%/3.57% loss up/down)
     bytes sent/received: 3360/3240
       send/receive rate: 480 bps / 462 bps
           packet length: 60 bytes
             timer stats: 0/56 (0.00%) missed, 0.01% error
 
X

xiaofan

High Supremacy Member
HWZ Insider
Joined
Sep 16, 2018
Messages
32,622
Reaction score
10,156
Last edited:
X

xiaofan

High Supremacy Member
HWZ Insider
Joined
Sep 16, 2018
Messages
32,622
Reaction score
10,156
X

xiaofan

High Supremacy Member
HWZ Insider
Joined
Sep 16, 2018
Messages
32,622
Reaction score
10,156
Testing of nttcp.

WiFi 6 TX test (upload) --> nttcp is a bit faster than iperf3, just as what CNX Software tested.

Wireless device: Acer Windows 11 laptop with Intel AX201 WiFi 6 adpter
Wireless router: Asus RT-AX86U, CH60, 160MHz channel bandwidth, 3m away
Receiver on Linux (LxC container running on Intel N100 based PVE 8.1 host)

Bash: 
root@ubuntu2204ct01:~# ntttcp -r -m 1,*,192.168.50.15 -t 60 -V
NTTTCP for Linux 1.4.0
---------------------------------------------------------
*** receiver role
ports:                           1
cpu affinity:                    *
server address:                  192.168.50.15
domain:                          IPv4
protocol:                        TCP
server port starting at:         5001
receiver socket buffer (bytes):  65536
test warm-up (sec):              no
test duration (sec):             60
test cool-down (sec):            no
show system tcp retransmit:      no
verbose mode:                    enabled
---------------------------------------------------------
13:21:19 DBG : user limits for maximum number of open files: soft: 1024; hard: 524288
13:21:19 INFO: 2 threads created
13:21:19 DBG : ntttcp server is listening on 192.168.50.15:5000

PS C:\work\speedtest\ntttcp> .\ntttcp.exe -s -m 1,*,192.168.50.15 -l 128K -t 60 -V -ns
Copyright Version 5.39
buffers_length: 131072
num_buffers_to_send: 9223372036854775807
send_socket_buff: -1
recv_socket_buff: -1
port: 5001
sync_port: 0
no_sync: 1
wait_timeout_milliseconds: 600000
async_flag: 0
verbose_flag: 1
wsa_flag: 0
use_ipv6_flag: 0
send_flag: 1
udp_flag: 0
udp_unconnected_flag: 0
verify_data_flag: 0
wait_all_flag: 0
run_time: 60000
warmup_time: 0
cooldown_time: 0
dash_n_timeout: 10800000
bind_sender_flag: 0
sender_name:
max_active_threads: 1
no_delay: 0
node_affinity: -1
udp_uso_size: 0
udp_receive_coalescing: 0
tp_flag: 0
use_hvsocket_flag: 0
no_stdio_buffer: 0
throughput_Bpms: 0
cpu_burn: 0
latency_measurement: 0
use_io_compl_ports: 0
cpu_from_idle_flag: 0
get_estats: 0
qos_flag: 0
packet_period: 0
jitter_measurement: 0
mapping[0]: 1
4/24/2024 21:25:51 proc_speed: 2419 MHz
4/24/2024 21:25:51 SetupThreads
4/24/2024 21:25:51 Threads: 1   Processor: -1   Host: 192.168.50.15
4/24/2024 21:25:51 created thread 0 port 5001
4/24/2024 21:25:51 StartSenderReceiver start thread 0 port 5001
4/24/2024 21:25:51 SetupNet port 5001
4/24/2024 21:25:51 connected to port 5001
4/24/2024 21:25:51 SetupNet complete on port 5001
4/24/2024 21:25:51 All threads ready!
4/24/2024 21:25:51 Network activity progressing...
4/24/2024 21:25:51 test start
4/24/2024 21:25:51 start recording results for sample 0
4/24/2024 21:26:51 stop recording results for sample 0
4/24/2024 21:26:51 test finish
4/24/2024 21:26:51 StartSenderReceiver done thread 0 port 5001
4/24/2024 21:26:51 PrintOutput


Thread  Time(s) Throughput(KB/s) Avg B / Compl
======  ======= ================ =============
     0   60.008       112676.177    131072.000


#####  Totals:  #####


   Bytes(MEG)    realtime(s) Avg Frame Size Throughput(MB/s)
================ =========== ============== ================
     6603.000000      60.003       1460.004          110.045


Throughput(Buffers/s) Cycles/Byte       Buffers
===================== =========== =============
              880.361       4.007     52824.000


DPCs(count/s) Pkts(num/DPC)   Intr(count/s) Pkts(num/intr)
============= ============= =============== ==============
     1185.930         2.739        4968.478          0.654


Packets Sent Packets Received Retransmits Errors Avg. CPU %
============ ================ =========== ====== ==========
     4742279           194899          41      0      2.389

Corresponding iperf3 test (Upload)

Bash: 
PS C:\work\speedtest\iperf-3.16-win64> .\iperf3.exe -c 192.168.50.15
Connecting to host 192.168.50.15, port 5201
[  5] local 192.168.50.175 port 3073 connected to 192.168.50.15 port 5201
[ ID] Interval           Transfer     Bitrate
[  5]   0.00-1.00   sec   106 MBytes   886 Mbits/sec
[  5]   1.00-2.01   sec   104 MBytes   873 Mbits/sec
[  5]   2.01-3.01   sec   110 MBytes   917 Mbits/sec
[  5]   3.01-4.00   sec   104 MBytes   881 Mbits/sec
[  5]   4.00-5.01   sec   100 MBytes   835 Mbits/sec
[  5]   5.01-6.01   sec  65.4 MBytes   546 Mbits/sec
[  5]   6.01-7.01   sec  98.6 MBytes   832 Mbits/sec
[  5]   7.01-8.00   sec   108 MBytes   908 Mbits/sec
[  5]   8.00-9.00   sec   105 MBytes   879 Mbits/sec
[  5]   9.00-10.01  sec  91.4 MBytes   762 Mbits/sec
- - - - - - - - - - - - - - - - - - - - - - - - -
[ ID] Interval           Transfer     Bitrate
[  5]   0.00-10.01  sec   992 MBytes   832 Mbits/sec                  sender
[  5]   0.00-10.05  sec   990 MBytes   826 Mbits/sec                  receiver

iperf Done.
 
Mach3.2

Mach3.2

Great Supremacy Member
HWZ Insider
Joined
Apr 8, 2011
Messages
72,405
Reaction score
2,466
Difference (if any) might be more obvious at 10GbE and higher.
I'm going to discount the wifi testing since a lot of factors can affect wifi results.

Perhaps the difference might be lower CPU overhead with ntttcp vs iperf3, since iperf3 is emulated using cygwin.
 
X

xiaofan

High Supremacy Member
HWZ Insider
Joined
Sep 16, 2018
Messages
32,622
Reaction score
10,156
Mach3.2 said:
Difference (if any) might be more obvious at 10GbE and higher.
I'm going to discount the wifi testing since a lot of factors can affect wifi results.

Perhaps the difference might be lower CPU overhead with ntttcp vs iperf3, since iperf3 is emulated using cygwin.
Click to expand...

Here is a "wired testing" with vitutal network adapters between two LxC containers on the PVE 8.1 host (Intel N100) since I do not have 10G gears.

Looks like ntttcp is a bit slower than iperf3 under Linux, but not too bad.

I cannot test Windows as I do not have Windows VMs installed.

Bash: 
root@debian12ct01:~# ntttcp -s -m 1,*,192.168.50.15 -l 128K -t 60
NTTTCP for Linux 1.4.0
---------------------------------------------------------
13:51:10 INFO: 1 threads created
13:51:10 INFO: 128 connections created in 3411 microseconds
13:51:10 INFO: Network activity progressing...
13:52:10 INFO: Test run completed.
13:52:10 INFO: Test cycle finished.
13:52:10 INFO: 128 connections tested
13:52:10 INFO: #####  Totals:  #####
13:52:10 INFO: test duration    :60.00 seconds
13:52:10 INFO: total bytes      :242279776256
13:52:10 INFO:   throughput     :32.30Gbps
13:52:10 INFO:   retrans segs   :143
13:52:10 INFO: cpu cores        :2
13:52:10 INFO:   cpu speed      :2899.981MHz
13:52:10 INFO:   user           :2.20%
13:52:10 INFO:   system         :36.27%
13:52:10 INFO:   idle           :43.52%
13:52:10 INFO:   iowait         :2.10%
13:52:10 INFO:   softirq        :15.90%
13:52:10 INFO:   cycles/byte    :0.81
13:52:10 INFO: cpu busy (all)   :100.93%
---------------------------------------------------------

root@debian12ct01:~# iperf3 -c 192.168.50.15
Connecting to host 192.168.50.15, port 5201
[  5] local 192.168.50.16 port 50690 connected to 192.168.50.15 port 5201
[ ID] Interval           Transfer     Bitrate         Retr  Cwnd
[  5]   0.00-1.00   sec  4.27 GBytes  36.7 Gbits/sec    0    535 KBytes
[  5]   1.00-2.00   sec  4.19 GBytes  36.0 Gbits/sec    0    632 KBytes
[  5]   2.00-3.00   sec  4.22 GBytes  36.3 Gbits/sec    0    847 KBytes
[  5]   3.00-4.00   sec  4.07 GBytes  35.0 Gbits/sec    0   1.14 MBytes
[  5]   4.00-5.00   sec  4.15 GBytes  35.7 Gbits/sec    0   1.26 MBytes
[  5]   5.00-6.00   sec  4.28 GBytes  36.8 Gbits/sec    0   1.26 MBytes
[  5]   6.00-7.00   sec  4.17 GBytes  35.8 Gbits/sec    0   1.26 MBytes
[  5]   7.00-8.00   sec  4.16 GBytes  35.8 Gbits/sec    0   1.53 MBytes
[  5]   8.00-9.00   sec  4.16 GBytes  35.7 Gbits/sec    0   1.53 MBytes
[  5]   9.00-10.00  sec  4.17 GBytes  35.8 Gbits/sec    0   1.53 MBytes
- - - - - - - - - - - - - - - - - - - - - - - - -
[ ID] Interval           Transfer     Bitrate         Retr
[  5]   0.00-10.00  sec  41.9 GBytes  36.0 Gbits/sec    0             sender
[  5]   0.00-10.04  sec  41.9 GBytes  35.8 Gbits/sec                  receiver

iperf Done.
 
X

xiaofan

High Supremacy Member
HWZ Insider
Joined
Sep 16, 2018
Messages
32,622
Reaction score
10,156
Another test with NAT.

Receiver -- LxC container 192.168.50.15
OpenWRT virtual router WAN: 192.168.50.168
OpenWRT virtual router LAN: 192.168.48.1
Sender -- LxC container 192.168.48.6

ntttcp is quite a bit slow compared to iperf3.

Bash: 
root@ubuntu2204ct21:~# ntttcp -s -m 1,*,192.168.50.15 -l 128K -t 60
NTTTCP for Linux 1.4.0
---------------------------------------------------------
14:03:26 INFO: Test cycle time negotiated is: 60 seconds
14:03:26 INFO: 1 threads created
14:03:26 INFO: 128 connections created in 10580 microseconds
14:03:26 INFO: Network activity progressing...
14:03:36 INFO: Test run completed.
14:03:36 INFO: Test cooldown is in progress...
14:04:26 INFO: Test cycle finished.
14:04:26 INFO: receiver exited from current test
14:04:26 INFO: 128 connections tested
14:04:26 INFO: #####  Totals:  #####
14:04:26 INFO: test duration    :10.31 seconds
14:04:26 INFO: total bytes      :17909415936
14:04:26 INFO:   throughput     :13.89Gbps
14:04:26 INFO:   retrans segs   :46631
14:04:26 INFO: cpu cores        :2
14:04:26 INFO:   cpu speed      :2899.808MHz
14:04:26 INFO:   user           :12.37%
14:04:26 INFO:   system         :64.22%
14:04:26 INFO:   idle           :13.55%
14:04:26 INFO:   iowait         :0.74%
14:04:26 INFO:   softirq        :9.12%
14:04:26 INFO:   cycles/byte    :2.89
14:04:26 INFO: cpu busy (all)   :53.48%
---------------------------------------------------------
root@ubuntu2204ct21:~# iperf3 -c 192.168.50.15
Connecting to host 192.168.50.15, port 5201
[  5] local 192.168.48.6 port 60664 connected to 192.168.50.15 port 5201
[ ID] Interval           Transfer     Bitrate         Retr  Cwnd
[  5]   0.00-1.00   sec  1.87 GBytes  16.0 Gbits/sec   32   2.87 MBytes       
[  5]   1.00-2.00   sec  1.92 GBytes  16.5 Gbits/sec    0   3.01 MBytes       
[  5]   2.00-3.00   sec  1.87 GBytes  16.0 Gbits/sec    0   3.01 MBytes       
[  5]   3.00-4.00   sec  1.86 GBytes  16.0 Gbits/sec   96   3.01 MBytes       
[  5]   4.00-5.00   sec  1.83 GBytes  15.8 Gbits/sec    0   3.01 MBytes       
[  5]   5.00-6.00   sec  1.97 GBytes  17.0 Gbits/sec    0   3.01 MBytes       
[  5]   6.00-7.00   sec  1.99 GBytes  17.1 Gbits/sec    0   3.01 MBytes       
[  5]   7.00-8.00   sec  2.00 GBytes  17.2 Gbits/sec    0   3.01 MBytes       
[  5]   8.00-9.00   sec  1.89 GBytes  16.3 Gbits/sec    0   3.01 MBytes       
[  5]   9.00-10.00  sec  1.92 GBytes  16.5 Gbits/sec    0   3.01 MBytes       
- - - - - - - - - - - - - - - - - - - - - - - - -
[ ID] Interval           Transfer     Bitrate         Retr
[  5]   0.00-10.00  sec  19.1 GBytes  16.4 Gbits/sec  128             sender
[  5]   0.00-10.04  sec  19.1 GBytes  16.4 Gbits/sec                  receiver

iperf Done.
 
You must log in or register to reply here.
Important Forum Advisory Note
This forum is moderated by volunteer moderators who will react only to members' feedback on posts. Moderators are not employees or representatives of HWZ Forums. Forum members and moderators are responsible for their own posts. Please refer to our Community Guidelines and Standards and Terms and Conditions for more information.
Top