Flent-users discussion archives
 help / color / mirror / Atom feed
* [Flent-users] [tohojo/flent] flent: extremely high ping times on macOS (#178)
@ 2019-09-06 11:57 Rich Brown
  2019-09-06 14:49 ` Dave Täht
                   ` (16 more replies)
  0 siblings, 17 replies; 18+ messages in thread
From: Rich Brown @ 2019-09-06 11:57 UTC (permalink / raw)
  To: tohojo/flent; +Cc: Subscribed


[-- Attachment #1.1: Type: text/plain, Size: 2353 bytes --]

Using Flent 1.3.0 with fping 4.2 on macOS 10.2.6:

I see high ping times (>100 msec in the following report) despite the fact that actual ping times (measured by fping during the flent test) are 20-30 msec, and none higher that 47 msec. (I am on 7mbps/768kbps DSL, so the 6.95mbps seems accurate.)

Here is the data file. [tcp_ndown-2019-09-06T074544.888052.Rich_Test.flent.gz](https://github.com/tohojo/flent/files/3583797/tcp_ndown-2019-09-06T074544.888052.Rich_Test.flent.gz)

What other information could I provide? Thanks.

```
bash-3.2$ flent -x -H flent-fremont.bufferbloat.net -t "Rich_Test" --te=download_streams=4 tcp_ndown
Started Flent 1.3.0 using Python 2.7.15.
Starting tcp_ndown test. Expected run time: 70 seconds.
Data file written to ./tcp_ndown-2019-09-06T074544.888052.Rich_Test.flent.gz.

Summary of tcp_ndown test run from 2019-09-06 11:45:44.888052
  Title: 'Rich_Test'

                             avg       median          # data pts
 Ping (ms) ICMP   :       104.03       105.00 ms              350
 TCP download avg :         1.71         1.74 Mbits/s         300
 TCP download sum :         6.83         6.95 Mbits/s         300
 TCP download::1  :         1.55         1.56 Mbits/s         281
 TCP download::2  :         1.85         1.83 Mbits/s         282
 TCP download::3  :         1.53         1.54 Mbits/s         282
 TCP download::4  :         1.91         1.90 Mbits/s         287
bash-3.2$
```

```
bash-3.2$ fping --version
fping: Version 4.2
fping: comments to david@schweikert.ch

bash-3.2$ fping -c 1000 gstatic.com -D
[1567770338.594778] gstatic.com : [0], 84 bytes, 20.7 ms (20.7 avg, 0% loss)
[1567770339.600662] gstatic.com : [1], 84 bytes, 21.7 ms (21.2 avg, 0% loss)
[1567770340.602339] gstatic.com : [2], 84 bytes, 21.5 ms (21.3 avg, 0% loss)
[1567770341.605205] gstatic.com : [3], 84 bytes, 22.2 ms (21.5 avg, 0% loss)
[1567770342.608899] gstatic.com : [4], 84 bytes, 21.0 ms (21.4 avg, 0% loss)
[1567770343.613397] gstatic.com : [5], 84 bytes, 21.4 ms (21.4 avg, 0% loss)
[1567770344.618612] gstatic.com : [6], 84 bytes, 21.5 ms (21.4 avg, 0% loss)
[1567770345.618789] gstatic.com : [7], 84 bytes, 20.8 ms (21.3 avg, 0% loss)
...
```

-- 
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/tohojo/flent/issues/178

[-- Attachment #1.2: Type: text/html, Size: 3691 bytes --]

[-- Attachment #2: Type: text/plain, Size: 151 bytes --]

_______________________________________________
Flent-users mailing list
Flent-users@flent.org
http://flent.org/mailman/listinfo/flent-users_flent.org

^ permalink raw reply	[flat|nested] 18+ messages in thread

* Re: [Flent-users] [tohojo/flent] flent: extremely high ping times on macOS (#178)
  2019-09-06 11:57 [Flent-users] [tohojo/flent] flent: extremely high ping times on macOS (#178) Rich Brown
@ 2019-09-06 14:49 ` Dave Täht
  2019-09-06 15:15 ` Rich Brown
                   ` (15 subsequent siblings)
  16 siblings, 0 replies; 18+ messages in thread
From: Dave Täht @ 2019-09-06 14:49 UTC (permalink / raw)
  To: tohojo/flent; +Cc: Subscribed


[-- Attachment #1.1: Type: text/plain, Size: 2732 bytes --]

Rich Brown <notifications@github.com> writes:\r
\r
> Using Flent 1.3.0 with fping 4.2 on macOS 10.2.6:\r
>\r
> I see high ping times (>100 msec in the following report) despite the\r
> fact that actual ping times (measured by fping during the flent test)\r
> are 20-30 msec, and none higher that 47 msec. (I am on 7mbps/768kbps\r
> DSL, so the 6.95mbps seems accurate.)\r
>\r
> Here is the data file.\r
> tcp_ndown-2019-09-06T074544.888052.Rich_Test.flent.gz\r
>\r
> What other information could I provide? Thanks.\r
>\r
> bash-3.2$ flent -x -H flent-fremont.bufferbloat.net -t "Rich_Test" --te=download_streams=4 tcp_ndown\r
> Started Flent 1.3.0 using Python 2.7.15.\r
> Starting tcp_ndown test. Expected run time: 70 seconds.\r
> Data file written to ./tcp_ndown-2019-09-06T074544.888052.Rich_Test.flent.gz.\r
>\r
> Summary of tcp_ndown test run from 2019-09-06 11:45:44.888052\r
>   Title: 'Rich_Test'\r
>\r
>                              avg       median          # data pts\r
>  Ping (ms) ICMP   :       104.03       105.00 ms              350\r
\r
I have seen this statistic be wildly wrong in the summary but correct in\r
the graph.\r
\r
>  TCP download avg :         1.71         1.74 Mbits/s         300\r
>  TCP download sum :         6.83         6.95 Mbits/s         300\r
>  TCP download::1  :         1.55         1.56 Mbits/s         281\r
>  TCP download::2  :         1.85         1.83 Mbits/s         282\r
>  TCP download::3  :         1.53         1.54 Mbits/s         282\r
>  TCP download::4  :         1.91         1.90 Mbits/s         287\r
> bash-3.2$\r
> bash-3.2$ fping --version\r
> fping: Version 4.2\r
> fping: comments to david@schweikert.ch\r
>\r
> bash-3.2$ fping -c 1000 gstatic.com -D\r
> [1567770338.594778] gstatic.com : [0], 84 bytes, 20.7 ms (20.7 avg, 0% loss)\r
> [1567770339.600662] gstatic.com : [1], 84 bytes, 21.7 ms (21.2 avg, 0% loss)\r
> [1567770340.602339] gstatic.com : [2], 84 bytes, 21.5 ms (21.3 avg, 0% loss)\r
> [1567770341.605205] gstatic.com : [3], 84 bytes, 22.2 ms (21.5 avg, 0% loss)\r
> [1567770342.608899] gstatic.com : [4], 84 bytes, 21.0 ms (21.4 avg, 0% loss)\r
> [1567770343.613397] gstatic.com : [5], 84 bytes, 21.4 ms (21.4 avg, 0% loss)\r
> [1567770344.618612] gstatic.com : [6], 84 bytes, 21.5 ms (21.4 avg, 0% loss)\r
> [1567770345.618789] gstatic.com : [7], 84 bytes, 20.8 ms (21.3 avg, 0% loss)\r
> ...\r
>\r
> —\r
> You are receiving this because you are subscribed to this thread.\r
> Reply to this email directly, view it on GitHub, or mute the thread.\r
\r
\r
-- \r
You are receiving this because you are subscribed to this thread.\r
Reply to this email directly or view it on GitHub:\r
https://github.com/tohojo/flent/issues/178#issuecomment-528885331

[-- Attachment #1.2: Type: text/html, Size: 4499 bytes --]

[-- Attachment #2: Type: text/plain, Size: 151 bytes --]

_______________________________________________
Flent-users mailing list
Flent-users@flent.org
http://flent.org/mailman/listinfo/flent-users_flent.org

^ permalink raw reply	[flat|nested] 18+ messages in thread

* Re: [Flent-users] [tohojo/flent] flent: extremely high ping times on macOS (#178)
  2019-09-06 11:57 [Flent-users] [tohojo/flent] flent: extremely high ping times on macOS (#178) Rich Brown
  2019-09-06 14:49 ` Dave Täht
@ 2019-09-06 15:15 ` Rich Brown
  2019-09-06 17:58 ` Toke Høiland-Jørgensen
                   ` (14 subsequent siblings)
  16 siblings, 0 replies; 18+ messages in thread
From: Rich Brown @ 2019-09-06 15:15 UTC (permalink / raw)
  To: tohojo/flent; +Cc: Subscribed


[-- Attachment #1.1: Type: text/plain, Size: 347 bytes --]

Hmmm, not this time... What else could I check?

![image](https://user-images.githubusercontent.com/1094930/64439228-7e5a5880-d097-11e9-9068-1f80fd634be7.png)


-- 
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/tohojo/flent/issues/178#issuecomment-528895661

[-- Attachment #1.2: Type: text/html, Size: 1910 bytes --]

[-- Attachment #2: Type: text/plain, Size: 151 bytes --]

_______________________________________________
Flent-users mailing list
Flent-users@flent.org
http://flent.org/mailman/listinfo/flent-users_flent.org

^ permalink raw reply	[flat|nested] 18+ messages in thread

* Re: [Flent-users] [tohojo/flent] flent: extremely high ping times on macOS (#178)
  2019-09-06 11:57 [Flent-users] [tohojo/flent] flent: extremely high ping times on macOS (#178) Rich Brown
  2019-09-06 14:49 ` Dave Täht
  2019-09-06 15:15 ` Rich Brown
@ 2019-09-06 17:58 ` Toke Høiland-Jørgensen
  2019-09-06 18:58 ` Rich Brown
                   ` (13 subsequent siblings)
  16 siblings, 0 replies; 18+ messages in thread
From: Toke Høiland-Jørgensen @ 2019-09-06 17:58 UTC (permalink / raw)
  To: tohojo/flent; +Cc: Subscribed


[-- Attachment #1.1: Type: text/plain, Size: 575 bytes --]

Rich Brown <notifications@github.com> writes:

> Hmmm, not this time... What else could I check?
>
> ![image](https://user-images.githubusercontent.com/1094930/64439228-7e5a5880-d097-11e9-9068-1f80fd634be7.png)

If you turn on debug logging during the test run (-L option), that log
will include the actual output of the ping tool being used, which you
can then compare to the graphs...


-- 
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/tohojo/flent/issues/178#issuecomment-528953835

[-- Attachment #1.2: Type: text/html, Size: 1983 bytes --]

[-- Attachment #2: Type: text/plain, Size: 151 bytes --]

_______________________________________________
Flent-users mailing list
Flent-users@flent.org
http://flent.org/mailman/listinfo/flent-users_flent.org

^ permalink raw reply	[flat|nested] 18+ messages in thread

* Re: [Flent-users] [tohojo/flent] flent: extremely high ping times on macOS (#178)
  2019-09-06 11:57 [Flent-users] [tohojo/flent] flent: extremely high ping times on macOS (#178) Rich Brown
                   ` (2 preceding siblings ...)
  2019-09-06 17:58 ` Toke Høiland-Jørgensen
@ 2019-09-06 18:58 ` Rich Brown
  2019-09-06 22:46 ` Toke Høiland-Jørgensen
                   ` (12 subsequent siblings)
  16 siblings, 0 replies; 18+ messages in thread
From: Rich Brown @ 2019-09-06 18:58 UTC (permalink / raw)
  To: tohojo/flent; +Cc: Subscribed


[-- Attachment #1.1: Type: text/plain, Size: 312 bytes --]

Here's the log file: https://pastebin.com/aBXCgjD1

NB I'll be away for the weekend, so won't follow up for a while. Thanks.

-- 
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/tohojo/flent/issues/178#issuecomment-528974379

[-- Attachment #1.2: Type: text/html, Size: 1741 bytes --]

[-- Attachment #2: Type: text/plain, Size: 151 bytes --]

_______________________________________________
Flent-users mailing list
Flent-users@flent.org
http://flent.org/mailman/listinfo/flent-users_flent.org

^ permalink raw reply	[flat|nested] 18+ messages in thread

* Re: [Flent-users] [tohojo/flent] flent: extremely high ping times on macOS (#178)
  2019-09-06 11:57 [Flent-users] [tohojo/flent] flent: extremely high ping times on macOS (#178) Rich Brown
                   ` (3 preceding siblings ...)
  2019-09-06 18:58 ` Rich Brown
@ 2019-09-06 22:46 ` Toke Høiland-Jørgensen
  2019-09-08 22:33 ` Rich Brown
                   ` (11 subsequent siblings)
  16 siblings, 0 replies; 18+ messages in thread
From: Toke Høiland-Jørgensen @ 2019-09-06 22:46 UTC (permalink / raw)
  To: tohojo/flent; +Cc: Subscribed


[-- Attachment #1.1: Type: text/plain, Size: 1391 bytes --]

Rich Brown <notifications@github.com> writes:

> Here's the log file: https://pastebin.com/aBXCgjD1
>
> NB I'll be away for the weekend, so won't follow up for a while.
> Thanks.

Looks like those are the actual ping times reported by fping:

[1567795827.309700] flent-fremont.bufferbloat.net : [175], 84 bytes, 95.5 ms (104 avg, 0% loss)
[1567795827.514875] flent-fremont.bufferbloat.net : [176], 84 bytes, 99.3 ms (104 avg, 0% loss)
[1567795827.715928] flent-fremont.bufferbloat.net : [177], 84 bytes, 100 ms (104 avg, 0% loss)
[1567795827.925782] flent-fremont.bufferbloat.net : [178], 84 bytes, 108 ms (104 avg, 0% loss)
[1567795828.129695] flent-fremont.bufferbloat.net : [179], 84 bytes, 111 ms (104 avg, 0% loss)
[1567795828.334842] flent-fremont.bufferbloat.net : [180], 84 bytes, 116 ms (104 avg, 0% loss)
[1567795828.528711] flent-fremont.bufferbloat.net : [181], 84 bytes, 110 ms (104 avg, 0% loss)
[1567795828.735846] flent-fremont.bufferbloat.net : [182], 84 bytes, 114 ms (105 avg, 0% loss)
[1567795828.918825] flent-fremont.bufferbloat.net : [183], 84 bytes, 96.8 ms (104 avg, 0% loss)
[1567795829.129416] flent-fremont.bufferbloat.net : [184], 84 bytes, 103 ms (104 avg, 0% loss)

-Toke


-- 
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/tohojo/flent/issues/178#issuecomment-529038202

[-- Attachment #1.2: Type: text/html, Size: 2862 bytes --]

[-- Attachment #2: Type: text/plain, Size: 151 bytes --]

_______________________________________________
Flent-users mailing list
Flent-users@flent.org
http://flent.org/mailman/listinfo/flent-users_flent.org

^ permalink raw reply	[flat|nested] 18+ messages in thread

* Re: [Flent-users] [tohojo/flent] flent: extremely high ping times on macOS (#178)
  2019-09-06 11:57 [Flent-users] [tohojo/flent] flent: extremely high ping times on macOS (#178) Rich Brown
                   ` (4 preceding siblings ...)
  2019-09-06 22:46 ` Toke Høiland-Jørgensen
@ 2019-09-08 22:33 ` Rich Brown
  2019-09-08 23:04 ` Toke Høiland-Jørgensen
                   ` (10 subsequent siblings)
  16 siblings, 0 replies; 18+ messages in thread
From: Rich Brown @ 2019-09-08 22:33 UTC (permalink / raw)
  To: tohojo/flent; +Cc: Subscribed


[-- Attachment #1.1: Type: text/plain, Size: 1239 bytes --]

Ahah! I see the difference. I was testing against `gstatic.com` with Terminal; Flent was testing against `flent-fremont.bufferbloat.net`

This seems a pretty big disparity between test results. Is there an option to tell Flent to ping a more local host? Thanks.

```
bash-3.2$ fping -c 1000 gstatic.com -D
[1567981240.102843] gstatic.com : [0], 84 bytes, 20.6 ms (20.6 avg, 0% loss)
[1567981241.103361] gstatic.com : [1], 84 bytes, 20.3 ms (20.4 avg, 0% loss)
[1567981242.107343] gstatic.com : [2], 84 bytes, 20.4 ms (20.4 avg, 0% loss)
^C
gstatic.com : xmt/rcv/%loss = 3/3/0%, min/avg/max = 20.3/20.4/20.6
bash-3.2$ fping -c 1000 flent-fremont.bufferbloat.net -D
[1567981266.195681] flent-fremont.bufferbloat.net : [0], 84 bytes, 92.9 ms (92.9 avg, 0% loss)
[1567981267.197573] flent-fremont.bufferbloat.net : [1], 84 bytes, 92.5 ms (92.7 avg, 0% loss)
[1567981268.199153] flent-fremont.bufferbloat.net : [2], 84 bytes, 92.9 ms (92.8 avg, 0% loss)
[1567981269.200396] flent-fremont.bufferbloat.net : [3], 84 bytes, 92.9 ms (92.8 avg, 0% loss)
...
```

-- 
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/tohojo/flent/issues/178#issuecomment-529247828

[-- Attachment #1.2: Type: text/html, Size: 2647 bytes --]

[-- Attachment #2: Type: text/plain, Size: 151 bytes --]

_______________________________________________
Flent-users mailing list
Flent-users@flent.org
http://flent.org/mailman/listinfo/flent-users_flent.org

^ permalink raw reply	[flat|nested] 18+ messages in thread

* Re: [Flent-users] [tohojo/flent] flent: extremely high ping times on macOS (#178)
  2019-09-06 11:57 [Flent-users] [tohojo/flent] flent: extremely high ping times on macOS (#178) Rich Brown
                   ` (5 preceding siblings ...)
  2019-09-08 22:33 ` Rich Brown
@ 2019-09-08 23:04 ` Toke Høiland-Jørgensen
  2019-09-09  0:06 ` Rich Brown
                   ` (9 subsequent siblings)
  16 siblings, 0 replies; 18+ messages in thread
From: Toke Høiland-Jørgensen @ 2019-09-08 23:04 UTC (permalink / raw)
  To: tohojo/flent; +Cc: Subscribed


[-- Attachment #1.1: Type: text/plain, Size: 637 bytes --]

Rich Brown <notifications@github.com> writes:

> Ahah! I see the difference. I was testing against `gstatic.com` with Terminal; Flent was testing against `flent-fremont.bufferbloat.net`
>
> This seems a pretty big disparity between test results. Is there an
> option to tell Flent to ping a more local host? Thanks.

You cannot (easily) override the existing ping host, but you can add an
additional one with --test-parameter ping_hosts=gstatic.com


-- 
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/tohojo/flent/issues/178#issuecomment-529250877

[-- Attachment #1.2: Type: text/html, Size: 2048 bytes --]

[-- Attachment #2: Type: text/plain, Size: 151 bytes --]

_______________________________________________
Flent-users mailing list
Flent-users@flent.org
http://flent.org/mailman/listinfo/flent-users_flent.org

^ permalink raw reply	[flat|nested] 18+ messages in thread

* Re: [Flent-users] [tohojo/flent] flent: extremely high ping times on macOS (#178)
  2019-09-06 11:57 [Flent-users] [tohojo/flent] flent: extremely high ping times on macOS (#178) Rich Brown
                   ` (6 preceding siblings ...)
  2019-09-08 23:04 ` Toke Høiland-Jørgensen
@ 2019-09-09  0:06 ` Rich Brown
  2019-09-09 12:25 ` Toke Høiland-Jørgensen
                   ` (8 subsequent siblings)
  16 siblings, 0 replies; 18+ messages in thread
From: Rich Brown @ 2019-09-09  0:06 UTC (permalink / raw)
  To: tohojo/flent; +Cc: Subscribed


[-- Attachment #1.1: Type: text/plain, Size: 3795 bytes --]

Hmmm... The `--test-parameter=ping_hosts=gstatic.com` information appears in the command-line summary but not in the charts. Is this correct?

If so, the charts display significantly different latency depending on which -H host gets used.  The Terminal display (bottom) shows fairly consistent ping times and transfer rate measurements using either of the two hosts.

This raises two questions:

1. Could flent charts use the `--test-parameter=ping_hosts=...` instead of values from the -H host? 

2. Why is there a discrepency between the `--test-parameter=ping_hosts=...` reading (showing 30-38 msec in my tests) and a separate, concurrent fping test (see original post, where it was running 20-22 msec)?

Thanks.

**Test to flent-fremont.bufferbloat.net gives latency > 100 msec:**
<img width="586" alt="image" src="https://user-images.githubusercontent.com/1094930/64496066-28d8a400-d26f-11e9-89ea-8b1a03b69872.png">

**Test to netperf.bufferbloat.net gives latency 55-65 msec:**
<img width="586" alt="image" src="https://user-images.githubusercontent.com/1094930/64496060-1d857880-d26f-11e9-9793-a36c7e0b0423.png">

**Text summary from flent-fremont.bufferbloat.net:**
```
bash-3.2$ flent -x -H flent-fremont.bufferbloat.net -t "Rich_Test" --test-parameter=ping_hosts=gstatic.com --te=download_streams=4 tcp_ndown
Started Flent 1.3.0 using Python 2.7.15.
Starting tcp_ndown test. Expected run time: 70 seconds.
Data file written to ./tcp_ndown-2019-09-08T193622.518606.Rich_Test.flent.gz.

Summary of tcp_ndown test run from 2019-09-08 23:36:22.518606
  Title: 'Rich_Test'

                                               avg       median          # data pts
 Ping (ms) ICMP                     :       104.28       105.00 ms              350
 Ping (ms) ICMP gstatic.com (extra) :        33.10        33.65 ms              350
 TCP download avg                   :         1.72         1.74 Mbits/s         301
 TCP download sum                   :         6.87         6.95 Mbits/s         301
 TCP download::1                    :         1.72         1.68 Mbits/s         285
 TCP download::2                    :         1.85         1.77 Mbits/s         287
 TCP download::3                    :         1.62         1.61 Mbits/s         284
 TCP download::4                    :         1.70         1.70 Mbits/s         283
bash-3.2$
```

**Text summary from netperf.bufferbloat.net:**
```
bash-3.2$ flent -x -H netperf.bufferbloat.net -t "netperf.bufferbloat.net" --test-parameter=ping_hosts=gstatic.com --te=download_streams=4 tcp_ndown
Started Flent 1.3.0 using Python 2.7.15.
Starting tcp_ndown test. Expected run time: 70 seconds.
Data file written to ./tcp_ndown-2019-09-08T194842.664611.netperf_bufferbloat_net.flent.gz.

Summary of tcp_ndown test run from 2019-09-08 23:48:42.664611
  Title: 'netperf.bufferbloat.net'

                                               avg       median          # data pts
 Ping (ms) ICMP                     :        54.51        56.30 ms              349
 Ping (ms) ICMP gstatic.com (extra) :        35.73        38.00 ms              349
 TCP download avg                   :         1.73         1.74 Mbits/s         300
 TCP download sum                   :         6.92         6.95 Mbits/s         300
 TCP download::1                    :         2.40         2.28 Mbits/s         279
 TCP download::2                    :         1.19         1.09 Mbits/s         259
 TCP download::3                    :         2.47         2.32 Mbits/s         278
 TCP download::4                    :         0.96         0.92 Mbits/s         253
bash-3.2$
```

-- 
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/tohojo/flent/issues/178#issuecomment-529258895

[-- Attachment #1.2: Type: text/html, Size: 5729 bytes --]

[-- Attachment #2: Type: text/plain, Size: 151 bytes --]

_______________________________________________
Flent-users mailing list
Flent-users@flent.org
http://flent.org/mailman/listinfo/flent-users_flent.org

^ permalink raw reply	[flat|nested] 18+ messages in thread

* Re: [Flent-users] [tohojo/flent] flent: extremely high ping times on macOS (#178)
  2019-09-06 11:57 [Flent-users] [tohojo/flent] flent: extremely high ping times on macOS (#178) Rich Brown
                   ` (7 preceding siblings ...)
  2019-09-09  0:06 ` Rich Brown
@ 2019-09-09 12:25 ` Toke Høiland-Jørgensen
  2019-09-09 15:07 ` Rich Brown
                   ` (7 subsequent siblings)
  16 siblings, 0 replies; 18+ messages in thread
From: Toke Høiland-Jørgensen @ 2019-09-09 12:25 UTC (permalink / raw)
  To: tohojo/flent; +Cc: Subscribed


[-- Attachment #1.1: Type: text/plain, Size: 1144 bytes --]

Rich Brown <notifications@github.com> writes:

> Hmmm... The `--test-parameter=ping_hosts=gstatic.com` information
> appears in the command-line summary but not in the charts. Is this
> correct?
>
> If so, the charts display significantly different latency depending on
> which -H host gets used. The Terminal display (bottom) shows fairly
> consistent ping times and transfer rate measurements using either of
> the two hosts.
>
> This raises two questions:
>
> 1. Could flent charts use the `--test-parameter=ping_hosts=...` instead of values from the -H host? 

Well, it should probably show both by default. For now, there's a
separate plot type you can use to see the extra flows (-p ping_extra).

> 2. Why is there a discrepency between the
> `--test-parameter=ping_hosts=...` reading (showing 30-38 msec in my
> tests) and a separate, concurrent fping test (see original post, where
> it was running 20-22 msec)?

Good question. IPv4 vs IPv6 maybe?


-- 
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/tohojo/flent/issues/178#issuecomment-529448466

[-- Attachment #1.2: Type: text/html, Size: 2655 bytes --]

[-- Attachment #2: Type: text/plain, Size: 151 bytes --]

_______________________________________________
Flent-users mailing list
Flent-users@flent.org
http://flent.org/mailman/listinfo/flent-users_flent.org

^ permalink raw reply	[flat|nested] 18+ messages in thread

* Re: [Flent-users] [tohojo/flent] flent: extremely high ping times on macOS (#178)
  2019-09-06 11:57 [Flent-users] [tohojo/flent] flent: extremely high ping times on macOS (#178) Rich Brown
                   ` (8 preceding siblings ...)
  2019-09-09 12:25 ` Toke Høiland-Jørgensen
@ 2019-09-09 15:07 ` Rich Brown
  2019-09-09 15:10 ` Rich Brown
                   ` (6 subsequent siblings)
  16 siblings, 0 replies; 18+ messages in thread
From: Rich Brown @ 2019-09-09 15:07 UTC (permalink / raw)
  To: tohojo/flent; +Cc: Subscribed


[-- Attachment #1.1: Type: text/plain, Size: 709 bytes --]

> > 1. Could flent charts use the `--test-parameter=ping_hosts=...` instead of values from the -H host?

> Well, it should probably show both by default. For now, there's a separate plot type you can use to see the extra flows (-p ping_extra).

I don't see how to use the `-p` option. When I use it in `flent --gui -p ping_extra ...filename.gz` I see the following message:

`ERROR: Plot 'ping_extra' failed: Error loading plotter: Unable to find plot configuration 'ping_extra'. Falling back to default plot 'totals'.` 


-- 
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/tohojo/flent/issues/178#issuecomment-529522201

[-- Attachment #1.2: Type: text/html, Size: 2218 bytes --]

[-- Attachment #2: Type: text/plain, Size: 151 bytes --]

_______________________________________________
Flent-users mailing list
Flent-users@flent.org
http://flent.org/mailman/listinfo/flent-users_flent.org

^ permalink raw reply	[flat|nested] 18+ messages in thread

* Re: [Flent-users] [tohojo/flent] flent: extremely high ping times on macOS (#178)
  2019-09-06 11:57 [Flent-users] [tohojo/flent] flent: extremely high ping times on macOS (#178) Rich Brown
                   ` (9 preceding siblings ...)
  2019-09-09 15:07 ` Rich Brown
@ 2019-09-09 15:10 ` Rich Brown
  2019-09-09 15:13 ` Rich Brown
                   ` (5 subsequent siblings)
  16 siblings, 0 replies; 18+ messages in thread
From: Rich Brown @ 2019-09-09 15:10 UTC (permalink / raw)
  To: tohojo/flent; +Cc: Subscribed


[-- Attachment #1.1: Type: text/plain, Size: 3514 bytes --]

> > 2. Why is there a discrepency between the
 `--test-parameter=ping_hosts=...` reading (showing 30-38 msec in my
 tests) and a separate, concurrent fping test (see original post, where
 it was running 20-22 msec)?

> Good question. IPv4 vs IPv6 maybe?

I don't think so. Here's fping for both atl.richb-hanover.com (IPv4 only) and atl6 (IPv6)...

```
bash-3.2$ fping -c 1000 atl.richb-hanover.com  -D
[1568033075.420184] atl.richb-hanover.com : [0], 84 bytes, 40.4 ms (40.4 avg, 0% loss)
[1568033076.423893] atl.richb-hanover.com : [1], 84 bytes, 40.9 ms (40.6 avg, 0% loss)
[1568033077.426595] atl.richb-hanover.com : [2], 84 bytes, 41.7 ms (41.0 avg, 0% loss)
[1568033078.430768] atl.richb-hanover.com : [3], 84 bytes, 40.8 ms (40.9 avg, 0% loss)
^C
bash-3.2$ fping -c 1000 atl6.richb-hanover.com  -D -6
[1568033063.252647] atl6.richb-hanover.com : [0], 64 bytes, 40.0 ms (40.0 avg, 0% loss)
[1568033064.254237] atl6.richb-hanover.com : [1], 64 bytes, 40.8 ms (40.4 avg, 0% loss)
[1568033065.257423] atl6.richb-hanover.com : [2], 64 bytes, 40.7 ms (40.5 avg, 0% loss)
[1568033066.261700] atl6.richb-hanover.com : [3], 64 bytes, 40.8 ms (40.6 avg, 0% loss)
^C
atl6.richb-hanover.com : xmt/rcv/%loss = 4/4/0%, min/avg/max = 40.0/40.6/40.8
```

And the results using atl.richb-hanover.com for the Flent: 54.77 msec during the test with 40-ish msec readings from fping and ping immediately thereafter.

```
bash-3.2$ flent -x -H atl.richb-hanover.com -t "atl.richb-hanover.com" --te=download_streams=4 tcp_ndown
Started Flent 1.3.0 using Python 2.7.15.
Starting tcp_ndown test. Expected run time: 70 seconds.
Data file written to ./tcp_ndown-2019-09-09T110116.988820.atl_richb-hanover_com.flent.gz.

Summary of tcp_ndown test run from 2019-09-09 15:01:16.988820
  Title: 'atl.richb-hanover.com'

                             avg       median          # data pts
 Ping (ms) ICMP   :        54.77        57.05 ms              350
 TCP download avg :         1.73         1.74 Mbits/s         300
 TCP download sum :         6.94         6.96 Mbits/s         300
 TCP download::1  :         2.09         2.02 Mbits/s         274
 TCP download::2  :         1.56         1.44 Mbits/s         268
 TCP download::3  :         0.99         0.90 Mbits/s         251
 TCP download::4  :         2.39         2.31 Mbits/s         278

bash-3.2$ fping -D atl.richb-hanover.com -c 4
[1568041403.319308] atl.richb-hanover.com : [0], 84 bytes, 52.5 ms (52.5 avg, 0% loss)
[1568041404.312934] atl.richb-hanover.com : [1], 84 bytes, 41.0 ms (46.7 avg, 0% loss)
[1568041405.313312] atl.richb-hanover.com : [2], 84 bytes, 40.7 ms (44.7 avg, 0% loss)
[1568041406.318130] atl.richb-hanover.com : [3], 84 bytes, 40.9 ms (43.8 avg, 0% loss)
atl.richb-hanover.com : xmt/rcv/%loss = 4/4/0%, min/avg/max = 40.7/43.8/52.5

bash-3.2$ ping atl.richb-hanover.com
PING atl.richb-hanover.com (23.226.232.80): 56 data bytes
64 bytes from 23.226.232.80: icmp_seq=0 ttl=51 time=40.722 ms
64 bytes from 23.226.232.80: icmp_seq=1 ttl=51 time=41.313 ms
64 bytes from 23.226.232.80: icmp_seq=2 ttl=51 time=41.022 ms
64 bytes from 23.226.232.80: icmp_seq=3 ttl=51 time=40.871 ms
^C
--- atl.richb-hanover.com ping statistics ---
4 packets transmitted, 4 packets received, 0.0% packet loss
round-trip min/avg/max/stddev = 40.722/40.982/41.313/0.219 ms

bash-3.2$```



-- 
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/tohojo/flent/issues/178#issuecomment-529523344

[-- Attachment #1.2: Type: text/html, Size: 5046 bytes --]

[-- Attachment #2: Type: text/plain, Size: 151 bytes --]

_______________________________________________
Flent-users mailing list
Flent-users@flent.org
http://flent.org/mailman/listinfo/flent-users_flent.org

^ permalink raw reply	[flat|nested] 18+ messages in thread

* Re: [Flent-users] [tohojo/flent] flent: extremely high ping times on macOS (#178)
  2019-09-06 11:57 [Flent-users] [tohojo/flent] flent: extremely high ping times on macOS (#178) Rich Brown
                   ` (10 preceding siblings ...)
  2019-09-09 15:10 ` Rich Brown
@ 2019-09-09 15:13 ` Rich Brown
  2019-09-09 17:00 ` Toke Høiland-Jørgensen
                   ` (4 subsequent siblings)
  16 siblings, 0 replies; 18+ messages in thread
From: Rich Brown @ 2019-09-09 15:13 UTC (permalink / raw)
  To: tohojo/flent; +Cc: Subscribed


[-- Attachment #1.1: Type: text/plain, Size: 468 bytes --]

PS Why am I perseverating about this now? I have taken Dave's script and packaged it for ease of use. See [Coffee Shop Bloat Script](https://github.com/richb-hanover/coffee-shop-bloat-test)

I can conceal a lot of the options into the script if I know what they are. Thanks again.

-- 
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/tohojo/flent/issues/178#issuecomment-529525038

[-- Attachment #1.2: Type: text/html, Size: 1849 bytes --]

[-- Attachment #2: Type: text/plain, Size: 151 bytes --]

_______________________________________________
Flent-users mailing list
Flent-users@flent.org
http://flent.org/mailman/listinfo/flent-users_flent.org

^ permalink raw reply	[flat|nested] 18+ messages in thread

* Re: [Flent-users] [tohojo/flent] flent: extremely high ping times on macOS (#178)
  2019-09-06 11:57 [Flent-users] [tohojo/flent] flent: extremely high ping times on macOS (#178) Rich Brown
                   ` (11 preceding siblings ...)
  2019-09-09 15:13 ` Rich Brown
@ 2019-09-09 17:00 ` Toke Høiland-Jørgensen
  2019-09-09 17:01 ` Toke Høiland-Jørgensen
                   ` (3 subsequent siblings)
  16 siblings, 0 replies; 18+ messages in thread
From: Toke Høiland-Jørgensen @ 2019-09-09 17:00 UTC (permalink / raw)
  To: tohojo/flent; +Cc: Subscribed


[-- Attachment #1.1: Type: text/plain, Size: 1666 bytes --]

Rich Brown <notifications@github.com> writes:

>> > 2. Why is there a discrepency between the
>  `--test-parameter=ping_hosts=...` reading (showing 30-38 msec in my
>  tests) and a separate, concurrent fping test (see original post, where
>  it was running 20-22 msec)?
>
>> Good question. IPv4 vs IPv6 maybe?
>
> I don't think so. Here's fping for both atl.richb-hanover.com (IPv4 only) and atl6 (IPv6)...
>
> ```
> bash-3.2$ fping -c 1000 atl.richb-hanover.com  -D
> [1568033075.420184] atl.richb-hanover.com : [0], 84 bytes, 40.4 ms (40.4 avg, 0% loss)
> [1568033076.423893] atl.richb-hanover.com : [1], 84 bytes, 40.9 ms (40.6 avg, 0% loss)
> [1568033077.426595] atl.richb-hanover.com : [2], 84 bytes, 41.7 ms (41.0 avg, 0% loss)
> [1568033078.430768] atl.richb-hanover.com : [3], 84 bytes, 40.8 ms (40.9 avg, 0% loss)
> ^C
> bash-3.2$ fping -c 1000 atl6.richb-hanover.com  -D -6
> [1568033063.252647] atl6.richb-hanover.com : [0], 64 bytes, 40.0 ms (40.0 avg, 0% loss)
> [1568033064.254237] atl6.richb-hanover.com : [1], 64 bytes, 40.8 ms (40.4 avg, 0% loss)
> [1568033065.257423] atl6.richb-hanover.com : [2], 64 bytes, 40.7 ms (40.5 avg, 0% loss)
> [1568033066.261700] atl6.richb-hanover.com : [3], 64 bytes, 40.8 ms (40.6 avg, 0% loss)
> ^C
> atl6.richb-hanover.com : xmt/rcv/%loss = 4/4/0%, min/avg/max = 40.0/40.6/40.8
> ```
>

You're not setting the period; does adding -p 200 make a difference? I
suppose fping could be self-congesting if your link is very slow...



-- 
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/tohojo/flent/issues/178#issuecomment-529572916

[-- Attachment #1.2: Type: text/html, Size: 3249 bytes --]

[-- Attachment #2: Type: text/plain, Size: 151 bytes --]

_______________________________________________
Flent-users mailing list
Flent-users@flent.org
http://flent.org/mailman/listinfo/flent-users_flent.org

^ permalink raw reply	[flat|nested] 18+ messages in thread

* Re: [Flent-users] [tohojo/flent] flent: extremely high ping times on macOS (#178)
  2019-09-06 11:57 [Flent-users] [tohojo/flent] flent: extremely high ping times on macOS (#178) Rich Brown
                   ` (12 preceding siblings ...)
  2019-09-09 17:00 ` Toke Høiland-Jørgensen
@ 2019-09-09 17:01 ` Toke Høiland-Jørgensen
  2019-09-11 16:31 ` Toke Høiland-Jørgensen
                   ` (2 subsequent siblings)
  16 siblings, 0 replies; 18+ messages in thread
From: Toke Høiland-Jørgensen @ 2019-09-09 17:01 UTC (permalink / raw)
  To: tohojo/flent; +Cc: Subscribed


[-- Attachment #1.1: Type: text/plain, Size: 734 bytes --]

Rich Brown <notifications@github.com> writes:

>> > 1. Could flent charts use the `--test-parameter=ping_hosts=...` instead of values from the -H host?
>
>> Well, it should probably show both by default. For now, there's a separate plot type you can use to see the extra flows (-p ping_extra).
>
> I don't see how to use the `-p` option. When I use it in `flent --gui
> -p ping_extra ...filename.gz` I see the following message:

From the GUI the ping_extra plot should be selectable in the plot
selection box once you've loaded the data file...


-- 
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/tohojo/flent/issues/178#issuecomment-529573115

[-- Attachment #1.2: Type: text/html, Size: 2180 bytes --]

[-- Attachment #2: Type: text/plain, Size: 151 bytes --]

_______________________________________________
Flent-users mailing list
Flent-users@flent.org
http://flent.org/mailman/listinfo/flent-users_flent.org

^ permalink raw reply	[flat|nested] 18+ messages in thread

* Re: [Flent-users] [tohojo/flent] flent: extremely high ping times on macOS (#178)
  2019-09-06 11:57 [Flent-users] [tohojo/flent] flent: extremely high ping times on macOS (#178) Rich Brown
                   ` (13 preceding siblings ...)
  2019-09-09 17:01 ` Toke Høiland-Jørgensen
@ 2019-09-11 16:31 ` Toke Høiland-Jørgensen
  2019-09-12 12:26 ` Rich Brown
  2019-09-12 12:26 ` Rich Brown
  16 siblings, 0 replies; 18+ messages in thread
From: Toke Høiland-Jørgensen @ 2019-09-11 16:31 UTC (permalink / raw)
  To: tohojo/flent; +Cc: Subscribed


[-- Attachment #1.1: Type: text/plain, Size: 804 bytes --]

Rich Brown <notifications@github.com> writes:

> Update: I'm trying to get repeatable test showing a difference of latency between the Flent ping_host values and a separate/concurrent fping run. (I saw it two days ago, but can't reproduce now... This is roundabout way to say that, for the moment, the Flent ping_host and separate fping runs are giving consistent values...)
>
>> You're not setting the period; does adding -p 200 make a difference? I
> suppose fping could be self-congesting if your link is very slow...
>
> I'm not sure how to use `-p 200` - is that a flent parameter? Thanks.

No, an fping one...


-- 
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/tohojo/flent/issues/178#issuecomment-530459962

[-- Attachment #1.2: Type: text/html, Size: 2244 bytes --]

[-- Attachment #2: Type: text/plain, Size: 151 bytes --]

_______________________________________________
Flent-users mailing list
Flent-users@flent.org
http://flent.org/mailman/listinfo/flent-users_flent.org

^ permalink raw reply	[flat|nested] 18+ messages in thread

* Re: [Flent-users] [tohojo/flent] flent: extremely high ping times on macOS (#178)
  2019-09-06 11:57 [Flent-users] [tohojo/flent] flent: extremely high ping times on macOS (#178) Rich Brown
                   ` (14 preceding siblings ...)
  2019-09-11 16:31 ` Toke Høiland-Jørgensen
@ 2019-09-12 12:26 ` Rich Brown
  2019-09-12 12:26 ` Rich Brown
  16 siblings, 0 replies; 18+ messages in thread
From: Rich Brown @ 2019-09-12 12:26 UTC (permalink / raw)
  To: tohojo/flent; +Cc: Subscribed


[-- Attachment #1.1: Type: text/plain, Size: 947 bytes --]

I am going to close this for now. Two reasons:

1. The primary reason for this report was that I misunderstood how Flent measures latency for its default plot. Currently, that plot shows latency to the -H host. As noted in https://github.com/tohojo/flent/issues/181#issuecomment-530500478, there could be an option to substitute a nearby host for default ping measurements. This would be great because the measured latency would likely have lower variance, and any induced bloat would be a larger percentage of the total.

2. I can no longer reproduce the difference between Flent's reported latency (say, to ping_host) and a separate fping session. My original measurement process might have been faulty - I can no longer say how I got those values.

Thanks.

-- 
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/tohojo/flent/issues/178#issuecomment-530801400

[-- Attachment #1.2: Type: text/html, Size: 2719 bytes --]

[-- Attachment #2: Type: text/plain, Size: 151 bytes --]

_______________________________________________
Flent-users mailing list
Flent-users@flent.org
http://flent.org/mailman/listinfo/flent-users_flent.org

^ permalink raw reply	[flat|nested] 18+ messages in thread

* Re: [Flent-users] [tohojo/flent] flent: extremely high ping times on macOS (#178)
  2019-09-06 11:57 [Flent-users] [tohojo/flent] flent: extremely high ping times on macOS (#178) Rich Brown
                   ` (15 preceding siblings ...)
  2019-09-12 12:26 ` Rich Brown
@ 2019-09-12 12:26 ` Rich Brown
  16 siblings, 0 replies; 18+ messages in thread
From: Rich Brown @ 2019-09-12 12:26 UTC (permalink / raw)
  To: tohojo/flent; +Cc: Subscribed


[-- Attachment #1.1: Type: text/plain, Size: 194 bytes --]

Closed #178.

-- 
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/tohojo/flent/issues/178#event-2628972439

[-- Attachment #1.2: Type: text/html, Size: 1949 bytes --]

[-- Attachment #2: Type: text/plain, Size: 151 bytes --]

_______________________________________________
Flent-users mailing list
Flent-users@flent.org
http://flent.org/mailman/listinfo/flent-users_flent.org

^ permalink raw reply	[flat|nested] 18+ messages in thread

end of thread, other threads:[~2019-09-12 12:27 UTC | newest]

Thread overview: 18+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2019-09-06 11:57 [Flent-users] [tohojo/flent] flent: extremely high ping times on macOS (#178) Rich Brown
2019-09-06 14:49 ` Dave Täht
2019-09-06 15:15 ` Rich Brown
2019-09-06 17:58 ` Toke Høiland-Jørgensen
2019-09-06 18:58 ` Rich Brown
2019-09-06 22:46 ` Toke Høiland-Jørgensen
2019-09-08 22:33 ` Rich Brown
2019-09-08 23:04 ` Toke Høiland-Jørgensen
2019-09-09  0:06 ` Rich Brown
2019-09-09 12:25 ` Toke Høiland-Jørgensen
2019-09-09 15:07 ` Rich Brown
2019-09-09 15:10 ` Rich Brown
2019-09-09 15:13 ` Rich Brown
2019-09-09 17:00 ` Toke Høiland-Jørgensen
2019-09-09 17:01 ` Toke Høiland-Jørgensen
2019-09-11 16:31 ` Toke Høiland-Jørgensen
2019-09-12 12:26 ` Rich Brown
2019-09-12 12:26 ` Rich Brown

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox