[Flent-users] Re: bursts test does not complete (WAS: ERROR: Runner Burst 0 failed check: No command set for SilentProcessRunner)

[Toke Høiland-Jørgensen]

ch  writes:

> Hi :wave:
>
> On 29/03/2023 19.43, ch wrote:
>> On 29/03/2023 19.42, ch wrote:
>>> But the flent test itself never terminate. The process keeps running 
>>> in the background. Resulting in the test results being saved to disk.
>> 
>> Resulting in the test results *not* being saved to disk.
>
> I fell like I am getting more gray hairs by the hours. 洛 I cannot 
> figure out why this happens.
>
> The burst test seems to run. Flent never "completes". Resulting in no 
> data output. :|

It's crashing in the parsing step, leading to a hang; this is visible on
stdout, so I guess your runner script hides that?

$ ./run-flent bursts -l 10
Starting Flent 2.1.1+git.6959463f using Python 3.10.10.
Starting bursts test. Expected run time: 10 seconds.
Exception in thread Thread-8 (_handle_results):
Traceback (most recent call last):
  File "/usr/lib/python3.10/threading.py", line 1016, in _bootstrap_inner
self.run()
  File "/usr/lib/python3.10/threading.py", line 953, in run
self._target(*self._args, **self._kwargs)
  File "/usr/lib/python3.10/multiprocessing/pool.py", line 595, in 
_handle_results
cache[job]._set(i, obj)
  File "/usr/lib/python3.10/multiprocessing/pool.py", line 779, in _set
self._callback(self._value)
  File "/home/alrua/source/flent/flent/runners.py", line 625, in recv_result
result, raw_values, metadata = res
TypeError: cannot unpack non-iterable NoneType object

Should be fixed in the latest master, please try that.

Also, BTW, not sure what your wrapper script does, but Flent has a
built-in batch run facility which may be helpful for such multi-run
tests...

-Toke
___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: ERROR: Runner Burst 0 failed check: No command set for SilentProcessRunner

[Toke Høiland-Jørgensen]

ch  writes:

> On 29/03/2023 12.10, Toke Høiland-Jørgensen wrote:
>>> to run. Resulting in the node generating traffic for the full pipe for
>>> hours on end. My observation is the test does not terminate for unknown
>>> reasons.
>> Hmm, did you pull the latest version of traffic-gen? The version on
>> github had a bug where it would just go into an infinite loop and never
>> exit (and not produce any traffic either). Fixed that yesterday, so make
>> sure you pull the fixed version...
>
> Hmm. Good question
>
>> # ls -la traffic-gen 
>> -rwxr-xr-x 1 root root 22832 Mar 28 20:04 traffic-gen
>
> I did compile the binary yesterday afternoon. So I can only assume it 
> contains the tree commits from yesterday.
>
>> # git log --oneline
>> b316813 (HEAD -> master, origin/master, origin/HEAD) Use getrandom() instead 
>> of rand()/srand()
>> eedf705 Fix infinite loop in scheduling logic
>> 5c1007e Support setting the TOS byte on transmitted packets.
>
> (The application is missing the option for outputting a version number 
> or git hash using e.g. -v, -V, or --version)
>
>> and not produce any traffic either
>
> It did produce *lots* of traffic. Had to terminate flent process to 
> terminate the child traffic-gen processes.
> Attached is the graph of the network traffic levels (in MB/s) as 
> reported by Proxmox.

Ah! Seems I introduced another infinite loop while fixing the previous
one; pushed another fix, please try that...

-Toke
___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: ERROR: Runner Burst 0 failed check: No command set for SilentProcessRunner

[Toke Høiland-Jørgensen]

ch  writes:

> Hi Toke,
>
> On 28/03/2023 16.01, Toke Høiland-Jørgensen wrote:
>> Ah, dug out some emails which set me on the right path. Seems to be this
>> utility which, apparently, I wrote myself a decade ago, and then,
>> evidently, forgot all about:
>> 
>> https://github.com/tohojo/traffic-gen
>> 
>> Seems to work well enough, so go ahead and try it, I guess? :)
>
> Installed the latest flent and traffic-gen from source yesterday after 
> our mail conversations.
>
> Ran into an issue with the bursts test, where the flent test continues 
> to run. Resulting in the node generating traffic for the full pipe for 
> hours on end. My observation is the test does not terminate for unknown 
> reasons.

Hmm, did you pull the latest version of traffic-gen? The version on
github had a bug where it would just go into an infinite loop and never
exit (and not produce any traffic either). Fixed that yesterday, so make
sure you pull the fixed version...

-Toke
___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: ERROR: Runner Burst 0 failed check: No command set for SilentProcessRunner

[Toke Høiland-Jørgensen]

ch  writes:

> [google foo]
>
> https://www.google.com/search?q=%22traffic-gen%22+flent
>
> The best that turned up are link to old 2016 posts,
>
> https://lists.bufferbloat.net/pipermail/make-wifi-fast/2016-March/000413.html
>
> http://lists.infradead.org/pipermail/ath10k/2016-March/007101.html
>
> https://www.spinics.net/lists/linux-wireless/msg148958.html
>
> [more google foo]
>
> Looking for the binary name together with Git gives me confusing answers.
>
> https://www.google.com/search?q=%22traffic-gen%22+%22git%22
>
> Maybe Dave knows which traffic-gen binary the bursts tests uses?

Ah, dug out some emails which set me on the right path. Seems to be this
utility which, apparently, I wrote myself a decade ago, and then,
evidently, forgot all about:

https://github.com/tohojo/traffic-gen

Seems to work well enough, so go ahead and try it, I guess? :)

-Toke
___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: ERROR: Runner Burst 0 failed check: No command set for SilentProcessRunner

[Toke Høiland-Jørgensen]

Chriztoffer  writes:

> flent-users,
>
> When I try to run the "bursts" test, I get an error message
> complaining there is no command set for class SilentProcessRunner.
> Which leaves me a bit clueless, after trying to dig something forth
> from the github issue tracker. And looking at the source code for the
> class. (What am I overlooking?)

Congratulations, you're the first person to try to use that test in a
least five years! (AFAICT it's been broken at least that long)

The immediate API issue you're seeing is fixable (pushed a fix to github
just now), but TBH I am not sure which utility that test is trying to
run. :/

It's calling a 'traffic-gen' utility, which is not really a googlable
name, and for the life of me I can't remember what that utility is
supposed to be. Do you have such a binary on your system?

-Toke
___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: flent <--> netperf on android?

[Toke Høiland-Jørgensen]

Erik Taraldsen  writes:

> Is there a netperf app that could be installed on android devices?  The
> search feature in the app store is, well not perfect so I struggle to find
> out for certain.

I'm not aware of any, no. You could probably cross-compile the binary to
the right architecture and just copy it over, but I really have very
little idea how the Android build system works...

-Toke
___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: Legend only once

[Toke Høiland-Jørgensen]

Erik Taraldsen  writes:

> I'm trying to automate testing and graphing with flent.  For my purposes it
> probably will make sense to use CDF and box plots.  However flent draws
> them differently.  In the CDF plot the legend is off to the right.  For box
> plot, the legend also is written vertically, and due to my long legends, is
> quite unreadable.  Fiddling with all legend options I can find on cli I'm
> not able to disable them.
>
> Anybody with a way to disable them?

Disabling the axis labels entirely is not possible (but you can remove
the regular legend with --no-legend). I usually use a combination of
--filter-legend --replace-legend, --filter-regexp, or straight up
--override-label to get sane axis markers for box plots (relying on just
the colours from the to-the-side legend quickly stops being legible once
you add more series to a box plot).

-Toke
___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: Aggregate test results

[Toke Høiland-Jørgensen]

Erik Taraldsen  writes:

> On Mon, Aug 8, 2022 at 1:55 PM Toke Høiland-Jørgensen  wrote:
>
>>
>> You mean, when loading data sets A and B (in separate .flent.gz files),
>> you want a graph of total throughput A+B? No, that's not currently
>> supported, you can only plot data files separately...
>>
>
> That is what I want to achieve, yes.  The intention is to have multiple
> clients running the same test at the same time.  Verify if the total
> capacity of the wifi router decreases under load or stays relatively the
> same.

So the way I usually do this is to drive all the client tests from a
single Flent instance. There are two ways to do this: Run Flent on the
server and connect to each client, running the test "in reverse", so to
speak - the --swap-up-down parameter is meant specifically for this.

The alternative is to use the remote execution features of Flent to
start the test binaries on the clients using ssh (works best using a
dedicated control network). This is what the --remote-host parameter is
for.

The benefit of driving everything in one test is that this way Flent
also takes care of synchronising the data series into a single run
(assuming clocks are reasonably synchronised across machines when using
--remote-host), and you'll get everything in one data file. It can be a
bit fiddly to set up, but you can encode everything into a single batch
file so repetitions are easy...

-Toke
___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: Aggregate test results

[Toke Høiland-Jørgensen]

Erik Taraldsen  writes:

> This may already be a feature of flent which I have yet to figure out.  If
> not, please consider this a feature request.
>
> In flent-gui, selecting multiple test files and viewing tcp upload/download
> - I would like to be able to have an aggregate throughput graph.

You mean, when loading data sets A and B (in separate .flent.gz files),
you want a graph of total throughput A+B? No, that's not currently
supported, you can only plot data files separately...

-Toke
___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: putting a floor under flent

[Toke Høiland-Jørgensen]

Dave Taht  writes:

> I am in the middle of writing a modest proposal for the NSF, to put
> more of an organisational floor under it, and am wondering if there
> are any other flent users out there, willing to submit a "letter of
> collaboration" as part of it?
>
> I'm also kind of curious how many hits per month flent.org gets?

Cloudflare says it has served 78.8k requests to 4.8k unique visitors in
the last 30 days; which is more than I would have guessed by a decent
margin.

> (hi toke, how big is this list?)

20, not including you and me and a couple of bots.

-Toke
___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: What other magic shims exist for pyqt?

[Toke Høiland-Jørgensen]

Dave Taht  writes:

>  "ERROR: Unable to find a usable Qt version."

You'll need 'apt install python3-pyside2 python3-matplotlib-qt5' as
well. See https://github.com/tohojo/flent/issues/259
___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] Added --test-payload option (PR #243)

[Toke Høiland-Jørgensen]

Merged #243 into master.

-- 
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/pull/243#event-5497802546___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] Added --test-payload option (PR #243)

[Toke Høiland-Jørgensen]

@tohojo requested changes on this pull request.

That's better! One more small nit below. Also, the commit message you added is 
great, but please squash the two commits together so that it becomes one commit 
that keeps the heading of the first one, and has the text of the second commit 
as its body.

Also, this sentence:

> Using the option will cause netperf to fail noisily, in case the specified 
> file is not readable or does not exist.

is slightly misleading. It's not netperf that fails noisily, it's the check in 
flent. So maybe "Flent will check if the file is readable before passing it to 
netperf, and fail the test run if it isn't" ?

>  except RunnerCheckError:
-netperf['buffer'] = ''
+if(fill_file == '/dev/urandom'):
+netperf['buffer'] = ''
+else:
+# If the custom file is not readable, fail noisily
+raise RunnerCheckError("The specified fill file does not 
exist or is not readable.")

The option is --test-payload, so referring to "fill file" in the error message 
is going to confuse people. Better stick to "test payload" here as well :)

-- 
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/pull/243#pullrequestreview-785005703___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] Added --test-payload option (PR #243)

[Toke Høiland-Jørgensen]

@tohojo requested changes on this pull request.

Please add a paragraph of text to the commit message describing why the feature 
is useful.

A few nits on the code, but otherwise looks good! :)

>  # Sanity check; is /dev/urandom readable? If so, use it to
 # pre-fill netperf's buffers
-self.run_simple(['dd', 'if=/dev/urandom', 'of=/dev/null', 
'bs=1', 'count=1'], errmsg="Err")
-netperf['buffer'] = '-F /dev/urandom'
+fillFile = args['test_payload']
+self.run_simple(['dd', f'if={fillFile}', 'of=/dev/null', 
'bs=1', 'count=1'], errmsg="Err")
+netperf['buffer'] = f'-F {fillFile}'

Flent runs on Python versions as old as 3.5, so we can't use f-strings, sadly...

> @@ -1022,10 +1023,14 @@ def check(self):
 netperf['-e'] = True
 
 try:
+# If --test-payload option is specified, use data from that 
file
+# else use the default value /dev/urandom. 
+# Below sanity check will be performed in either case.

I'm not sure if the sanity check makes sense for a user-specified file: it 
means we'll just silently ignore the option if the file is not readable. I 
think it's better to fail noisily. We could do this either by doing the check 
but erroring out if it fails (it's the default), or we could skip the test for 
custom values and just let netperf fail when it can't open it.

> @@ -378,6 +378,12 @@ def __call__(self, parser, namespace, values, 
> option_string=None):
 "to decrease the socket buffer size. Can be specified multiple times, "
 "with each value corresponding to a stream of a test.")
 
+test_group.add_argument(
+"--test-payload",
+action="store", type=unicode, dest="TEST_PAYLOAD", default='/dev/urandom',
+help="Path to file containing payload to pre-fill the netperf buffers with"

Missing space at the end of the first string here

>  # Sanity check; is /dev/urandom readable? If so, use it to
 # pre-fill netperf's buffers
-self.run_simple(['dd', 'if=/dev/urandom', 'of=/dev/null', 
'bs=1', 'count=1'], errmsg="Err")
-netperf['buffer'] = '-F /dev/urandom'
+fillFile = args['test_payload']

Variables are generally lowercase and underscore-separated rather then 
camelCase...

-- 
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/pull/243#pullrequestreview-783839744___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] Runners: stop stats collection iterators after test-time is completed (#237)

[Toke Høiland-Jørgensen]

@tohojo commented on this pull request.



> @@ -41,7 +41,8 @@ then
 fi
 
 command_string=$(cat 

[Flent-users] Re: [tohojo/flent] Runners: stop stats collection iterators after test-time is completed (#237)

[Toke Høiland-Jørgensen]

@tohojo requested changes on this pull request.

Unfortunately this approach doesn't work on openwrt; but then the existing 
timestamped output doesn't either, we need to use the tc_iterate C program on 
there anyway. So maybe that's okay, and since the binary uses a timerfd, maybe 
we don't even need to switch that over to using 'length' either (since the 
'count' should be accurate enough when the kernel runs the timer).

Some comments on the code below...

>  }
 
 while getopts "c:I:H:t:p:f:" opt; do
 case $opt in
-c) count="$OPTARG" ;;
+l) length="$OPTARG" ;;

You also need to change the opts string passed to getopts (`c:` to `l:`)

> @@ -10,6 +12,7 @@ while getopts "i:c:I:C:H:" opt; do
 case $opt in
 i) interface=$OPTARG ;;
 c) count=$OPTARG ;;
+l) length=$OPTARG ;;

Same thing as above re: changing the optstring, also here you're not removing 
the `count` option...

> @@ -1,7 +1,9 @@
 #!/bin/bash
+# set -x

Leftover debug setting?

> @@ -41,7 +41,8 @@ then
 fi
 
 command_string=$(cat < @@ -22,7 +25,8 @@ buffer=""
 
 command_string=$(cat  @@ -10,6 +12,7 @@ while getopts "i:c:I:C:H:" opt; do
 case $opt in
 i) interface=$OPTARG ;;
 c) count=$OPTARG ;;
+l) length=$OPTARG ;;

Ah, you're still passing `-c` to the C binary. Guess that should be changed as 
well to support the `-l` option. Which means we'll need to detect if it 
understands it; ugh...

-- 
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/pull/237#pullrequestreview-778537436___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] TCP (IPv4) connection fails in rrul46compete when specifying --local-bind (#242)

[Toke Høiland-Jørgensen]

Closed #242 via 5527b187bd4b8c9b3f6d919891f0b97a89edf000.

-- 
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/242#event-5395660127___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] -6 Option ignored for hosts with IPv4 and IPv6 address (#241)

[Toke Høiland-Jørgensen]

Closed #241 via 76c96e66f1dfb75bfa2c90f8aea414b91a911b68.

-- 
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/241#event-5394030111___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] -6 Option ignored for hosts with IPv4 and IPv6 address (#241)

[Toke Høiland-Jørgensen]

j-breyer ***@***.***> writes:

>>I also wonder what happens if you don't use --local-bind.
>
> irtt connection no longer fails, as there's no mismatch in address
> families. However, as long as `mn.h2` has an IPv4 defined, it uses
> IPv4 to connect, disregarding the `-6` param

Hmm, I guess maybe the Go runtime has its own idea as to what
constitutes the right way to lookup a hostname, then? I'll add the -6
arg for irtt, then...


-- 
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/241#issuecomment-932171051___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] -6 Option ignored for hosts with IPv4 and IPv6 address (#241)

[Toke Høiland-Jørgensen]

Hmm, could you please post a full debug log (from --log-file), and the
output of this command (with both address families defined in your /etc/hosts):

`python -c 'import socket; print(socket.getaddrinfo("mn.h1", None, 
socket.AF_UNSPEC, socket.SOCK_STREAM))'`


-- 
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/241#issuecomment-931373985___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] Questions (and Error) regarding rrul_prio (#240)

[Toke Høiland-Jørgensen]

Closed #240.

-- 
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/240#event-5388045656___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] Questions (and Error) regarding rrul_prio (#240)

[Toke Høiland-Jørgensen]

Awesome! Closing this issue, then :)

-- 
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/240#issuecomment-931360519___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] runners: Bracket IPv6 addresses before passing to irtt (#239)

[Toke Høiland-Jørgensen]

Awesome - 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/pull/239#issuecomment-931348708___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] irtt fails with IPv6 (wrong format) (#238)

[Toke Høiland-Jørgensen]

Closed #238 via 187dbb56ac308272d19cf4359ef85717fb7693ae.

-- 
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/238#event-5387927729___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] runners: Bracket IPv6 addresses before passing to irtt (#239)

[Toke Høiland-Jørgensen]

Merged #239 into master.

-- 
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/pull/239#event-5387927802___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] Questions (and Error) regarding rrul_prio (#240)

[Toke Høiland-Jørgensen]

j-breyer ***@***.***> writes:

> Well, now I get a
> ```
> UDP RTT test: Using irtt
> which: /usr/local/sbin/fping6 is not an executable file
> which: /usr/local/bin/fping6 is not an executable file
> which: /usr/sbin/fping6 is not an executable file
> which: Found fping6 executable at /usr/bin/fping6
> which: /usr/local/sbin/ping6 is not an executable file
> which: /usr/local/bin/ping6 is not an executable file
> which: /usr/sbin/ping6 is not an executable file
> which: Found ping6 executable at /usr/bin/ping6
> WARNING: Found fping, but couldn't parse its output. Not using.
> Looking up hostname 'mn.h1'.
> ERROR: Runner Ping (ms) ICMP failed check: Cannot parse output of the system 
> ping binary (/usr/bin/ping6). Please install fping v3.5+.
> ```
>
> `sudo apt install fping` tells me, `fping` is at version 4.2-1

Hmm, that's odd. What's the output of `fping6 -D -c 1 localhost` ?


-- 
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/240#issuecomment-931256026___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] Questions (and Error) regarding rrul_prio (#240)

[Toke Høiland-Jørgensen]

j-breyer ***@***.***> writes:

>>Distro packaging tends to lack behind a bit - which distro (and version)
> are you running?
>
> ```
> DISTRIB_ID=Ubuntu
> DISTRIB_RELEASE=20.04
> DISTRIB_CODENAME=focal
> DISTRIB_DESCRIPTION="Ubuntu 20.04.3 LTS"
> NAME="Ubuntu"
> VERSION="20.04.3 LTS (Focal Fossa)"
> ID=ubuntu
> ID_LIKE=debian
> ```

Ah, in that case there's a PPA you can use:
https://launchpad.net/~tohojo/+archive/ubuntu/flent


-- 
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/240#issuecomment-931232898___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] runners: Bracket IPv6 addresses before passing to irtt (#239)

[Toke Høiland-Jørgensen]

@tohojo approved this pull request.

LGTM, but could you please squash the changes into a single commit? :)



-- 
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/pull/239#pullrequestreview-767718801___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] irtt fails with IPv6 (wrong format) (#238)

[Toke Høiland-Jørgensen]

j-breyer ***@***.***> writes:

> I found another workaround by setting up non-scoped IPv6 addresses on the 
> hosts (e.g., `a::601` and `a::e02`).
> Then I added them to `/etc/hosts` with a related hostname, e.g., `a::601 
> mn.h1` and `a::e02 mn.h2`
>
> Running now `sudo flent rrul -p all_scaled -6 --local-bind mn.h1 -l 60 -H 
> mn.h2 -o rrul_hosts6.pdf -v` gives Error free output.
>
> However, I noticed a different bug when using this:
> If a host contains an IPv4 as well as an IPv6 address in `/etc/hosts`, flent 
> seems to always prefer the IPv4 entry. This results in the process either 
> failing, because it tries an irtt connection from IPv4 to IPv6 address, or 
> the above test runs the irtt connection completely in IPv4, ignoring the `-6` 
> argument, if both devices can be resolved to an IPv4
>
> Maybe this is content for a separate Issue?

Yes, please open a separate issue. Flent uses the host name resolution
mechanism, so in the absence of -4/-6, it should prefer whatever the
host does. However, a -6 should override this, so if it doesn't that's a
bug...


-- 
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/238#issuecomment-930581977___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] runners: Bracket IPv6 addresses before passing to irtt (#239)

[Toke Høiland-Jørgensen]

@tohojo requested changes on this pull request.



> @@ -1682,7 +1683,13 @@ def check(self):
 if self.ip_version is not None:
 args.append("-{}".format(self.ip_version))
 
-args.append(self.host)
+try:
+ipaddress.IPv6Address(self.host)
+host = "[{}]".format(self.host)
+except ValueError:
+host = self.host

Using the ipaddress module for parsing is fine, but let's move this into the 
normalise_host() utility function. Simply adding a second argument 'bracket_v6' 
which defaults to 'False' to that should do the trick :)

-- 
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/pull/239#pullrequestreview-767221497___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] Installed fping 4.2 but still observe error "Please install fping v3.5+ during flent runs" (#235)

[Toke Høiland-Jørgensen]

Closed #235.

-- 
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/235#event-5374101466___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] Installed fping 4.2 but still observe error "Please install fping v3.5+ during flent runs" (#235)

[Toke Høiland-Jørgensen]

See https://github.com/tohojo/flent/issues/232 - closing this as a duplicate

-- 
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/235#issuecomment-929246945___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] irtt fails with IPv6 (wrong format) (#238)

[Toke Høiland-Jørgensen]

Could you please include the full command you're running flent with, and the 
full log output? I'm not quite sure I understand what you're trying to do and 
what goes wrong from your description...

-- 
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/238#issuecomment-928275998___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] ss_iterate.sh unnecessarily runs for a long time (#236)

[Toke Høiland-Jørgensen]

Shashank D ***@***.***> writes:

>> Did you test whether this works on OpenWrt and on Dash?
> No. I could only test this on a Ubuntu machine.

Right, okay. Well, the 'dash' shell should be available to install on
ubuntu as well. Feel free to open a PR with the change, I can test it on
OpenWrt...


-- 
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/236#issuecomment-919500031___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] ss_iterate.sh unnecessarily runs for a long time (#236)

[Toke Høiland-Jørgensen]

Shashank D ***@***.***> writes:

>> Hmm, good question. So you're just running ss in a busy loop, or did you 
>> keep the 'sleep'? If so, that seems a bit excessive to do by default. 
> Yes I did keep the sleep in the loop. Here's the full loop:
> ```bash
> command_string=$(cat < duration="$(echo "$count*$interval" | bc) sec";
> endtime=\$(date -d "\$duration" +%s%N);
> while (( \$(date +%s%N) <= \$endtime )); do
> ss -t -i -p -n state connected "dst $target $filter"
> echo ''
> date '+Time: %s.%N';
> echo "---";
> sleep $interval || exit 1;
> done
> EOF
> )
> ```

Right, makes sense. Feel free to open a PR with this, but please lose
the dependency on 'bc'. Either do the math in native shell, or just pass
the duration as a parameter from Flent.

Did you test whether this works on OpenWrt and on Dash?


-- 
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/236#issuecomment-919035245___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] ss_iterate.sh unnecessarily runs for a long time (#236)

[Toke Høiland-Jørgensen]

Hmm, good question. So you're just running ss in a busy loop, or did you
keep the 'sleep'? If so, that seems a bit excessive to do by default.

I'd be a bit surprised if you could get this short a polling time out of
the shell script, actually; what interval were you able to achieve?


-- 
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/236#issuecomment-918112784___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] Installed fping 4.2 but still observe error "Please install fping v3.5+ during flent runs" (#235)

[Toke Høiland-Jørgensen]

Could you please post the verbose output (with `-v`)? That should shed some 
light on why it can't find your fping binary...

Also, what OS are you running this on?

-- 
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/235#issuecomment-909628349___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] runners: Add regex to parse FQ-PIE qdisc stats (#234)

[Toke Høiland-Jørgensen]

Merged #234 into master.

-- 
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/pull/234#event-5205745124___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] doesn't run on windows (#233)

[Toke Høiland-Jørgensen]

That particular error may be fixable, but there will be others: running tests 
on windows is not supported, only viewing existing data files.

You should be able to run tests from inside a WSL environment, though. See 
https://docs.microsoft.com/en-us/windows/wsl/about

-- 
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/233#issuecomment-901780482___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] ss parsing problem with ipv6 (#229)

[Toke Høiland-Jørgensen]

Closed #229 via e58e8de81fcffbbf0892954e45d673ab0d5bc4e2.

-- 
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/229#event-4943824368___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] [tohojo/flent] Release v2.0.1 - Flent v2.0.1

[Toke Høiland-Jørgensen]

Released on 2021-06-24.

Changes since v2.0.0 include:

- Fix globbing in plot and test definitions so it works on newer Python
  versions

- Don't hang forever while trying to process invalid time series data, and
  check fping timestamp output before using it (fixes a hang bug when using
  certain versions of fping on BSD and OSX)

- Don't crash on metadata collection if hexdump is not available on the system.

- Clarify that the Flent license doesn't apply to its output

-- 
You are receiving this because you are subscribed to this thread.
View it on GitHub:
https://github.com/tohojo/flent/releases/tag/v2.0.1___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Announcing Flent v2.0.1

[Toke Høiland-Jørgensen]

Hi everyone

This is to announce the release of Flent v2.0.1.

This is a small bugfix release fixing a few annoying bugs affecting
plotting on newer Python versions, and running tests on OSX with buggy
fping.

Changes since v2.0.0 include:

- Fix globbing in plot and test definitions so it works on newer Python
  versions

- Don't hang forever while trying to process invalid time series data, and
  check fping timestamp output before using it (fixes a hang bug when using
  certain versions of fping on BSD and OSX)

- Don't crash on metadata collection if hexdump is not available on the system.

- Clarify that the Flent license doesn't apply to its output

As always, the Flent source is available via PyPi. The sha256 checksums
for the source code archives are:

60d32d1409377b416836c01a35dce6c8ab83e8c83b3e9cba3f0563f5b5145b14  
flent-2.0.1-py2.py3-none-any.whl
300a09938dc2b4a0463c9144626f25e0bd736fd47806a9444719fa024d671796  
flent-2.0.1.tar.gz

Distro packages should already be updated, with the exception of the
PPA, which is still building, but should be there shortly.

-Toke


signature.asc
Description: PGP signature
___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] Unable to run Flent test on MacMini/MacOS - Test is not terminating (#228)

[Toke Høiland-Jørgensen]

crazyrai ***@***.***> writes:

> Flent test will not terminate and need Ctrl+c to terminate. Looks like
> underneath tests are running good, mostly unable to collect results
> and in loop.
>
> I tried on MacOS-Bigsur, Catalina, Mohave

This is a bug in fping - see #221. Please try the git version of Flent,
that should refuse to use the broken fping binary...


-- 
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/228#issuecomment-867560881___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] Plots are broken somehow (#227)

[Toke Høiland-Jørgensen]

"Kirill Lukonin (EvilWirelessMan)" ***@***.***> writes:

> @tohojo 
>
> Than you very much. Master version works perfectly.
> Is it a good approach to make a minor release 2.0.1 without this
> issue?

Yeah, I guess a 2.0.1 release would be in order :)


-- 
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/227#issuecomment-867542993___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] Plots are broken somehow (#227)

[Toke Høiland-Jørgensen]

Could you please test if this issue is still present in the git version?
You could just do:

git clone https://github.com/tohojo/flent
cd flent
./run-flent --input some_test_file.flent --plot totals --bounds-x 300 
--figure-width 26 --figure-height 14 --zero-y --no-title --bounds-y=500 
--output ./test_totals.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/227#issuecomment-865079245___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] seeing the max, rather than the average, on rrul is less distracting (#225)

[Toke Høiland-Jørgensen]

Problem with just using the max is those outliers (like that last datapoint in 
TCP upload on the second plot). I suppose adding a variant of the ping plot 
where the bold black line rides on top of the plot instead of being an average? 
Is that what you mean?

-- 
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/225#issuecomment-859521484___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] Flent in Docker? (#220)

[Toke Høiland-Jørgensen]

xciter327 ***@***.***> writes:

> Here is what I setup for a quick n dirty testing. It's a pretty "fat"
> image, but I don't feel like optimizing it at this point. You can
> probably do away with a bunch of dependencies if You relegate yourself
> to cli only Flent. It was based on a previous attempt I saw somewhere
> here.

Thanks! I'd be happy to include this as an example in the repository
under packaging/ if you'd care to open a pull request for it? :)


-- 
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/220#issuecomment-857077117___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] Issues running flent / packaging into container (#226)

[Toke Høiland-Jørgensen]

Closed #226 via 28ed6ebf349c1abf04179481aebea914b14d8d87.

-- 
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/226#event-4857606181___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] Issues running flent / packaging into container (#226)

[Toke Høiland-Jørgensen]

xciter327 ***@***.***> writes:

> I wanted to run some test on Fedora 34, which seems to have lost the netperf 
> package. So I decided to give packaging Flent into a container a go. I'm 
> having some issues actually running the thing.
> Output from tring to run it:
> ```
> ➜  flent git:(master) ✗ podman run -it --network=host 0c4ea2cce3e flent rrul 
> -p all_scaled -l 60 -H netperf.bufferbloat.net
> Started Flent 2.0.0 using Python 3.9.5.
> Traceback (most recent call last):
>   File "/usr/bin/flent", line 33, in 
> sys.exit(load_entry_point('flent==2.0.0', 'console_scripts', 'run')())
>   File "/usr/share/flent/flent/__init__.py", line 59, in run_flent
> b.run()
>   File "/usr/share/flent/flent/batch.py", line 617, in run
> return self.run_test(self.settings, self.settings.DATA_DIR, True)
>   File "/usr/share/flent/flent/batch.py", line 505, in run_test
> record_metadata(res, settings.EXTENDED_METADATA,
>   File "/usr/share/flent/flent/metadata.py", line 132, in record_metadata
> m['MODULE_VERSIONS'] = get_module_versions()
>   File "/usr/share/flent/flent/metadata.py", line 514, in get_module_versions
> version_strings.split(
> AttributeError: 'NoneType' object has no attribute 'split'
> ```

Hmm, my guess would be that this happens because the container doesn't
have 'hexdump' available. Really shouldn't cause a crash, though, that's
definitely a bug...

And yeah, I really ought to do something about that missing netperf
package (you're the second person pointing it out in the last couple of
days)...


-- 
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/226#issuecomment-856657814___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: centos 8 and cake and flent

[Toke Høiland-Jørgensen]

Dave Taht  writes:

> Currently centos (and I assume redhat) is at 4.18. Cake went into 4.19
> so I assume the next major
> redhat/centos releases will have it.

Note that the RHEL kernel version number is basically a complete
fabrication; it's the version that the kernel was forked from, something
like 30% of all commits are backported for each new upstream release,
without changing the RHEL-kernel version number.

Which means that all the Cake out-of-tree kernel version compatibility
stuff is not going to work, because that works based on the kernel
version number...

> Is there a yum/rpm expert in the house? flent does not appear to be
> packaged up for this (?),

It's in Fedora: https://src.fedoraproject.org/rpms/flent - should be
fairly straight-forward to add it to EPEL as well, but thus far no one
has requested it... :)

> neither is netperf or irtt. Is there a repo I could use?

netperf is blocked on licensing: 
https://bugzilla.redhat.com/show_bug.cgi?id=1729939

Since the re-licensing there may be a chance, but not sure what the
procedure is when there's not a release with the new license.

As for irtt, that should be pretty straight-forward to package.

-Toke
___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] add test conf files to test results (#224)

[Toke Høiland-Jørgensen]

moeller0 ***@***.***> writes:

> Hi Toke,
>
> I just got hold of a flent.gz file from somebody with a new test, and
> my flent can not open that file (makes sense, since the tests is
> unknown). Would it be an option to have the actual test.conf file
> embedded into the flent data, so opening somebody else's test would
> work? It would be very convenient, if not only opening of such files
> would work, but also extraction of the test file to the flent/test
> (unless the test exosts already). But maybe the whole idea has privacy
> issues that doom it?

Yeah, test.conf files are Python code; so no, that should definitely not
be embedded into a data file ;)

The right way to get around this would to make Flent smarter about
parsing out data series and plotting them without having the pre-defined
plots...


-- 
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/224#issuecomment-831205743___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] Add Gentoo support to your home page. (#223)

[Toke Høiland-Jørgensen]

Closed #223 via 94013620052d9a00da949b64550d7804d580e14e.

-- 
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/223#event-4647542029___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] Add Gentoo support to your home page. (#223)

[Toke Høiland-Jørgensen]

Awesome, 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/223#issuecomment-827166180___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] Add Gentoo support to your home page. (#223)

[Toke Høiland-Jørgensen]

"Gary E. Miller" ***@***.***> writes:

> Gentoo has an ebuild for flent: net-analyzer/flent
>
> Please add to your home page with the other distros.

Sure, absolutely! Thank you for the suggestion!

What's the command to install ebuilds on Gentoo these days? Just `emerge 
net-analyzer/flent` ?


-- 
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/223#issuecomment-827140889___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] flent on openbsd loops, consuming available cpu/memory, after test completes (#221)

[Toke Høiland-Jørgensen]

Closed #221 via 15908e9351fe87cf5a16e3ef0123dfb8ec6355e8.

-- 
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/221#event-4485789029___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] --remote-host doesn't seem to work correctly (#222)

[Toke Høiland-Jørgensen]

You're welcome! :)


-- 
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/222#issuecomment-800451421___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] --remote-host doesn't seem to work correctly (#222)

[Toke Høiland-Jørgensen]

bkiziuk ***@***.***> writes:

> I'm trying to use flent for running traffic between different host.
> According to what I have read and understand, the following command: 
> ```
> ./run-flent --host 192.168.98.5 tcp_2up  --remote-host ***@***.*** 
> --remote-host ***@***.*** --log-file=log.txt
> ```

The tcp_2up test has a ping flow, and two runners computing average and
totals. Those take up indexes 0, 1 and 2, so the actual TCP flows start
from 3. You can see this by running the test and looking at the metadata
(with `flent -f metadata mytestfile.flent.gz`) - look for the IDX value
in the objects in SERIES_META.



-- 
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/222#issuecomment-799830227___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] flent on openbsd loops, consuming available cpu/memory, after test completes (#221)

[Toke Høiland-Jørgensen]

Ah, nice find, thanks! Still, Flent should probably detect this and bail - 
going to reopen this issue to track that...

-- 
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/221#issuecomment-792346198___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] flent on openbsd loops, consuming available cpu/memory, after test completes (#221)

[Toke Høiland-Jørgensen]

Reopened #221.

-- 
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/221#event-4418443887___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] flent on openbsd loops, consuming available cpu/memory, after test completes (#221)

[Toke Høiland-Jørgensen]

Hmm, I've had a report of something that kinda sounds similar on OSX. Maybe 
they are related? Could you please try the latest git version? That adds a bit 
of extra debugging, which may help me correlate the two issues.

Also, something seems to be off with that line number in the last trace (there 
is no line 808 in aggregators.py) - any chance you can figure out the real 
place it's spinning?

How much memory is it allocating before it crashes / before you kill it, and 
how long does it spin for?

-- 
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/221#issuecomment-792123269___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] Flent in Docker? (#220)

[Toke Høiland-Jørgensen]

Rich Brown  writes:

> I read an intriguing article - [Run more stuff in
> Docker](https://jonathan.bergknoff.com/journal/run-more-stuff-in-docker/).
> It showed a solution to a big frustration: getting the right version
> of Flent, Python, netperf, and all the other libraries onto my
> computer without screwing up other software.
>
> Is anyone working on a Dockerfile that bundles up Flent with its
> preferred support packages? This sounds like a good enhancement to the
> already enhanced 2.0 version. I don't have a lot of time, but might
> take it on in the next month to six weeks.

I'm not a huge fan of Docker, so no, no plans for this. If you want to
take a stab at it, go ahead. It does feel a bit overkill for something
as small as Flent, though. What exactly is your problem with getting
Flent to run today?


-- 
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/220#issuecomment-768263305___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] support for flowgrind (#219)

[Toke Høiland-Jørgensen]

Jonas Smedegaard  writes:

> the tool flowgrind seems like it might be an interesting backend for
> flent to support: https://github.com/flowgrind/flowgrind
>
> From its README:
>
>> Unlike most cross-platform testing tools, flowgrind can output some
>> transport layer information, which are usually internal to the TCP/IP
>> stack. For example, on Linux and FreeBSD this includes among others
>> the kernel's estimation of the end-to-end RTT, the size of the TCP
>> congestion window (CWND) and slow start threshold (SSTHRESH).

These values Flent can already extract (using the 'ss' tool). Just
enable --socket-stats :)

Also from its README:

>  In contrast to similar tools like iperf or netperf it features a
>  distributed architecture, where throughput and other metrics are
>  measured between arbitrary flowgrind server processes.

...which makes me suspect it's not a very good fit for Flent.


-- 
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/219#issuecomment-766769369___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: flent: No valid data

[Toke Høiland-Jørgensen]

 writes:

> Dear All
>
> Some time ago, I stood up a Linux Mint VM on my windows box in which to run
> flent. Flent showed (of course) that I had the usual bloat.
>
> Since then, I have replaced my router with an Ubiquiti ER-4 so I could shape
> traffic with CAKE, getting unadulterated traffic from my VDSL modem  via IP
> passthrough.
>
> I have pre-CAKE flent data using this router. But I can't get data after
> starting CAKE.
>
>
> The issue is that I am getting "No valid Data" from flent.

Here's a hint:

Stdout: establish control: are you sure there is a netserver listening on
23.226.232.80 at port 12865?

You didn't include the command-line you passed to Flent, so can't see
which hostname it's trying to run against; but are you sure there's a
valid server running there? :)

-Toke
___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] [tohojo/flent] Release v2.0.0 - Flent v2.0.0

[Toke Høiland-Jørgensen]

Released on 2021-01-14.

This release drops support for Python 2 and Qt4, and associated older versions
of matplotlib. The major version bump is to signify this change; otherwise
consider this a (long overdue) regular incremental release over v1.3.0.

Changes since v1.3.0 include:

- Drop support for Python 2; the minimum required Python version is now 3.5.

- Drop support for old versions of libraries. The minimum supported version of
  matplotlib is 1.5, and for the GUI, Qt 5 is required, using either PyQt or
  PySide2. A new dependency on the 'QtPy' package is added to support both Qt
  bindings.

- Add parsing of WiFi device information as extended metadata (from Emilly
  Albuquerque).

- Add the ping_markings test parameter for setting DSCP values on ping flows.

- Add new plot types to rrul_be, tcp_nup and related tests (from Pete Heist).

- Add multiple-inheritance support for batch files (from Pete Heist).

- Support running multiple instances of http-getter for HTTP tests, by supplying
  the --http-getter-urllist option multiple times.

- Add new rrul_var test where the number of bidirectional streams can be
  configured via the bidir_streams test parameter.

- Support output of aggregate statistics per series also in csv format (via the
  'stats_csv' formatter).

- Make sure all underlying tools use the same values for symbolic diffserv
  markings by parsing the symbolic names into numeric values before passing them
  down to the tools.

- Support specifying custom symbolic diffserv markings using the --marking-name
  options. This can be used to specify site-specific names for diffserv markings
  (e.g., 'gaming') that can then be used when running a test.

- Support specifying the same hostname multiple times (for tests that take
  multiple targets). Previously, this could only be achieved by using different
  hostnames that all resolve to the same IP address; now, Flent will accept the
  same hostname multiple times, but will turn append a number after two slashes
  ('//N') to the display of the hostname when doing so.

- Support ranges and wildcards in the --remote-hosts parameter for running test
  runners remotely via SSH. For instance '--remote-host=1-3=host1' will run the
  first three runners on 'host1', or '--remote-host=*=host2' will run all
  runners on host2. If both concrete numbers (or ranges) and the wildcard are
  specified, the concrete numbers will take precedence.

- Fix several bugs related to plotting, the loading of old data files and RC
  files, running of commands etc.


-- 
You are receiving this because you are subscribed to this thread.
View it on GitHub:
https://github.com/tohojo/flent/releases/tag/v2.0.0___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Announcing Flent v2.0.0

[Toke Høiland-Jørgensen]

Hi everyone

This is to announce the release of Flent v2.0.0.

This release drops support for Python 2 and Qt4, and associated older versions
of matplotlib. The major version bump is to signify this change; otherwise
consider this a (long overdue) regular incremental release over v1.3.0.

Changes since v1.3.0 include:

- Drop support for Python 2; the minimum required Python version is now 3.5.

- Drop support for old versions of libraries. The minimum supported version of
  matplotlib is 1.5, and for the GUI, Qt 5 is required, using either PyQt or
  PySide2. A new dependency on the 'QtPy' package is added to support both Qt
  bindings.

- Add parsing of WiFi device information as extended metadata (from Emilly
  Albuquerque).

- Add the ping_markings test parameter for setting DSCP values on ping flows.

- Add new plot types to rrul_be, tcp_nup and related tests (from Pete Heist).

- Add multiple-inheritance support for batch files (from Pete Heist).

- Support running multiple instances of http-getter for HTTP tests, by supplying
  the --http-getter-urllist option multiple times.

- Add new rrul_var test where the number of bidirectional streams can be
  configured via the bidir_streams test parameter.

- Support output of aggregate statistics per series also in csv format (via the
  'stats_csv' formatter).

- Make sure all underlying tools use the same values for symbolic diffserv
  markings by parsing the symbolic names into numeric values before passing them
  down to the tools.

- Support specifying custom symbolic diffserv markings using the --marking-name
  options. This can be used to specify site-specific names for diffserv markings
  (e.g., 'gaming') that can then be used when running a test.

- Support specifying the same hostname multiple times (for tests that take
  multiple targets). Previously, this could only be achieved by using different
  hostnames that all resolve to the same IP address; now, Flent will accept the
  same hostname multiple times, but will turn append a number after two slashes
  ('//N') to the display of the hostname when doing so.

- Support ranges and wildcards in the --remote-hosts parameter for running test
  runners remotely via SSH. For instance '--remote-host=1-3=host1' will run the
  first three runners on 'host1', or '--remote-host=*=host2' will run all
  runners on host2. If both concrete numbers (or ranges) and the wildcard are
  specified, the concrete numbers will take precedence.

- Fix several bugs related to plotting, the loading of old data files and RC
  files, running of commands etc.

As always, the Flent source is available via PyPi. The sha256 checksums
for the source code archives are:

978a6e9a49b0236df32ee9e45e0a4dc7b172b812b7a6960ba6534f36f3d011ee  
flent-2.0.0-py2.py3-none-any.whl
8a9c6f828b4e8621c59ae74e28c33b501a5ba074470041ff6aa897c15ce9  
flent-2.0.0.tar.gz

The PPA should already be updated, as is the Arch package and the Fedora
packages. The native Debian package is pending but should appear as soon
as the Debian build infrastructure gets around to it.

-Toke


signature.asc
Description: PGP signature
___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] rrul tests are not being able to saturate link (#218)

[Toke Høiland-Jørgensen]

Closed #218.

-- 
You are receiving this because you commented.
Reply to this email directly or view it on GitHub:
https://github.com/tohojo/flent/issues/218#event-4202370704___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] Aborting plotting due to error (#187)

[Toke Høiland-Jørgensen]

Closed #187.

-- 
You are receiving this because you commented.
Reply to this email directly or view it on GitHub:
https://github.com/tohojo/flent/issues/187#event-4201992783___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] Question about custom symbolic diffserv marking names feature (#215)

[Toke Høiland-Jørgensen]

Closed #215.

-- 
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/215#event-4201977585___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] rrul tests are not being able to saturate link (#218)

[Toke Høiland-Jørgensen]

olg33  writes:

> In a related issue, It took our attention that whenever we run a rrul or 
> rrul_var test, upload traffic seems to be absent.
>
> Below examples were performed with two servers connected back-to-back
> with no router in between to rule out any blockage (they connect using
> an ethernet switch with ports in the same Vlan).

Erm, that does sound odd! Only obvious thing I can immediately think of
is maybe the network card is running in half-duplex mode? Did you check
the output of ethtool for the interface you're using?

Otherwise, to debug further you can try running with -L and looking at
the netperf output in the log file to see if there's any obvious
pointers there...


-- 
You are receiving this because you commented.
Reply to this email directly or view it on GitHub:
https://github.com/tohojo/flent/issues/218#issuecomment-747763365___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] How to create two simultaneous tests with different bandwidth in both hosts? (#217)

[Toke Høiland-Jørgensen]

lealog  writes:

>> Well, it would require an underlying test tool that could produce
>> this sort of traffic. I wouldn't be opposed to adding support for
>> such a tool if it materialised, but I am not aware of any either.
>
> I would say that we can do this using iperf3. Iperf3 can specify the
> DS and US bandwidth using the option "-b". Does this make sense for
> you?

The trouble with the iperf UDP tests it that they just send off packets
and don't provide any feedback on how much makes it through. You may or
may not get a result at the end from the other side at the end of the
test, but that can also easily get lost. So while it can generate a base
load, this kind of test is not really suitable for graphing data *about*
those flows during the test.

>> Some orchestration is possible already by a combination of tests with
>> multiple remote endpoints, and remote test runners. It requires a bit
>> of fiddling with the CLI and/or batch files to set up, but it's
>> doable. Extending support for running tests through the GUI would be
>> neat, but unless someone dedicates resources to this I don't think
>> it's likely to happen anytime soon, unfortunately...
>
> I would say that supporting this on GUI is a "nice to have" :)
> Regarding the CLI, how can we do this? Can you share your idea?

Well, there are two options: Reverse the setup and run the Flent test
from the "other side" with different flows going to each of your two
test hosts. Or run Flent on one of the two "local" hosts, and use the
--remote-host option to execute some of the test runners on the other
host (preferably using a dedicated control connection between the two).
I've been using the former setup for WiFi tests, but which one is
easiest to do really depends on the details of your test network...


-- 
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/217#issuecomment-747477879___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] rrul tests are not being able to saturate link (#218)

[Toke Høiland-Jørgensen]

olg33  writes:

> Hi,
>
> We are installing a new satellite link and we are using flent to run
> bandwidth testing. The idea is to saturate the link with marked and
> unmarked packets.
>
> However initial results show that the test was not able to reach the
> maximum bandwidth downstream supported by the link which is about
> 60Mbps.
>  
> For unmarked traffic we ran the following test:
> `flent rrul -H 10.111.40.252` 
>
> For marked traffic we used the following test:
> `flent rrul_var -H 10.111.40.252 --test-parameter bidir_streams=5 
> --test-parameter markings=11,13,15,19,21 `
>
> On both cases we reached no more than 50% of the link capacity.
>
> To rule out any problem with the link itself, we ran the following test:
> `flent tcp_12down -H 10.111.40.252`
>
> This time. we were able to reach saturation levels (60Mbps) of
> downstream traffic. Also ran iperf3 and obtained the same results
>
> Do you see any reason why the rrul and rrul_var tests aren't being
> able to generate enough traffic to saturate the link? Maybe any
> parameter I'm missing or something? Or maybe that's normal, anyhow,
> I'd appreciate any comment on this matter.

Hmm, when you say satellite link, I assume this has really high RTT,
right? This usually makes it really difficult for TCP to saturate the
connection; I believe providers try to improve on this with various
kinds of "accelerators" that mess with the TCP connection to try to
alleviate this problem. That could be failing when there's bidirectional
traffic?

Another thing to note is that when you're running bidirectional traffic
the bulk traffic will be competing with ACKs in each direction. In
particular, if there's a lot of queueing delay in the upstream, that
will delay the ACKs which can prevent TCP from ramping up properly on
the downstream. What kind of queueing is on the bottleneck link, and are
you seeing the latency increase?

By my guess the latter effect would be most likely, or maybe a
combination? You may be able to see something interesting if you capture
the traffic and look at some tcptrace plots of the transfers.

The RRUL test is deliberately designed to stress connections to induce
these sorts of weird failure cases, so I guess you could say it's not
unexpected. But obviously it's not an ideal functioning of the link :)


-- 
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/218#issuecomment-745233711___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] Question about custom symbolic diffserv marking names feature (#215)

[Toke Høiland-Jørgensen]

olg33  writes:

> Oops! Right I was forgetting the --test-parameter sentence. I tested again 
> and now all looks good.
>
> As always, thanks for your help.

Awesome - you're welcome! :)


-- 
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/215#issuecomment-741961660___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] Question about custom symbolic diffserv marking names feature (#215)

[Toke Høiland-Jørgensen]

olg33  writes:

> Done! This time all works just fine. :-)
>
> Just one more thing. I'd like to have the marking names declared in the 
> `.flentrc` file so that I don't have to type them every time I run a test. 
> With that in mind, I created the ~/.flentrc file and added the following 
> lines:
>   
> [global]
> `marking_names=10=0x28;11=0x2C;13=0x34;15=0x3C;17=0x44;19=0x4C;21=0x54;23=0x5C`
>
> But when I run a test:
> `flent rrul_var -H 10.111.38.60 --test-parameter bidir_streams=2 
> markings=11,13`
>
> The system returns the following error:
> `flent: error: unrecognized arguments: markings=11,13`

You have to repeat --test-parameter for each parameter. So try running:

flent rrul_var -H 10.111.38.60 --test-parameter bidir_streams=2 
--test-parameter markings=11,13

You can also double-check the loading of the RC file by running with -v;
that should make Flent print out whether it finds the file and which
values it loads from it...


-- 
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/215#issuecomment-741159062___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] support (re-)enabling netperf routing of UDP packets (#214)

[Toke Høiland-Jørgensen]

Closed #214.

-- 
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/214#event-4086781836___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] support (re-)enabling netperf routing of UDP packets (#214)

[Toke Høiland-Jørgensen]

Great! Closing this, then :)

-- 
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/214#issuecomment-741039524___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] How to create two simultaneous tests with different bandwidth in both hosts? (#217)

[Toke Høiland-Jørgensen]

lealog  writes:

> The goal is to evaluate and compare different Routers and QOS policies.
> Any plans to have Flent supporting this type of use cases? Also, could
> be interesting to have Flent managing different hosts and orchestrate
> them using a common GUI or CLI.

Well, it would require an underlying test tool that could produce this
sort of traffic. I wouldn't be opposed to adding support for such a tool
if it materialised, but I am not aware of any either.

> Unfortunately, I'm not a developer and I'm not able to provide code
> for some of these features.

Some orchestration is possible already by a combination of tests with
multiple remote endpoints, and remote test runners. It requires a bit of
fiddling with the CLI and/or batch files to set up, but it's doable.
Extending support for running tests through the GUI would be neat, but
unless someone dedicates resources to this I don't think it's likely to
happen anytime soon, unfortunately...


-- 
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/217#issuecomment-740752354___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] support (re-)enabling netperf routing of UDP packets (#214)

[Toke Høiland-Jørgensen]

Jonas Smedegaard  writes:

> Quoting Toke Høiland-Jørgensen (2020-12-06 13:03:02)
>> Or it may be that the same UDP issue you linked also affects the UDP_RR
>> test; I'll dig around in the netperf source code a bit and see what I
>> can find :)
>
> Thanks - much appreciated!

OK, poked around in the netperf sources a bit, and it seems the UDP_RR
test is *not* affected by the -R flag, so the issue is most likely with
some other source of blocking.

You could try setting up irtt on both server and client and see if that
fares better - Flent should automatically pick it up if it's installed
on the client, and you just need to start it along with netserver on the
server-side:

https://github.com/heistp/irtt/


-- 
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/214#issuecomment-740729292___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] How to create two simultaneous tests using flent-gui? (#216)

[Toke Høiland-Jørgensen]

lealog  writes:

> Thanks :)
> Are you planning to change the Flent version? I'm asking this because
> I want to test this "new" feature :)

Yeah, a new release is way overdue. Will try to get around to it before
the holidays :)


-- 
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/216#issuecomment-740723535___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] How to create two simultaneous tests using flent-gui? (#216)

[Toke Høiland-Jørgensen]

Closed #216 via bfea1ce694af66cac021cdd82093d056a6a039d7.

-- 
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/216#event-4084914339___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] How to create two simultaneous tests using flent-gui? (#216)

[Toke Høiland-Jørgensen]

lealog  writes:

> Hi,
>
> I'm trying to create two simultaneous tests using flent-gui but I'm
> always receiving the same error "RuntimeError: Found no hostnames on
> lookup of 192.168.56.103,192.168.56.102".
>
> Using the CLI I'm able to run the tests using "udo flent
> rtt_fair_var_mixed -l 60 -H 192.168.56.102 -H 192.168.56.103 -t
> test_two_hosts_6".
>
> Can you please clarify is this is supported by flent-gui?

Hmm, no, it isn't, but let's fix that :)


-- 
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/216#issuecomment-740635621___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] Question about custom symbolic diffserv marking names feature (#215)

[Toke Høiland-Jørgensen]

olg33  writes:

> I did clone flent:
>
> `git clone https://github.com/tohojo/flent`
>
> Then installed it:
>
> `sudo python3 setup.py install`
>
> It is he following release:
>
> `Started Flent 1.9.9-git-0024eb9 using Python 3.6.9.`
>
> Hopefully this is the right one.
>
> However, when I run flent with any test, I get the following error:

Ah, that's a bug! Seems I accidentally broke support for old versions of
Python, my apologies. Pushed a fix, so try doing another 'git pull' and
see if that helps :)


-- 
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/215#issuecomment-740610277___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] Question about custom symbolic diffserv marking names feature (#215)

[Toke Høiland-Jørgensen]

olg33  writes:

> This one:
>
> `Started Flent 1.9.9-git-b8165da using Python 3.6.9.`

That's from before the --marking-name parameter was introduced, which
would explain why it doesn't work :)

Try doing a 'git pull'...


-- 
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/215#issuecomment-740212063___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] Question about custom symbolic diffserv marking names feature (#215)

[Toke Høiland-Jørgensen]

olg33  writes:

> Thanks Toke, In order to do further testing with your suggestions, first I 
> got rid of the .flentrc file (in part because it seemed not to be properly 
> read by flent) and inserted the values directly in the command line using the 
> --marking-name parameter described in cf4f3da:
>
> `flent rrul_var -H 10.111.38.60 --test-parameter bidir_streams=2 
> --marking-name 11=0x2C --marking-name 13=0x34 --test-parameter markings=11,13`
>
> However, the parameter is not being recognized:
>
>  `flent: error: unrecognized arguments: --marking-name 11=0x2C --marking-name 
> 13=0x34`
>
> is my syntax correct?

Yeah, should be; what version of Flent are you using?


-- 
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/215#issuecomment-740085368___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] Question about custom symbolic diffserv marking names feature (#215)

[Toke Høiland-Jørgensen]

olg33  writes:

> Hello,
>
> I have a system using the feature that implements custom symbolic diffserv 
> marking names, In the past, I got it working with your help (Issue #203).
>
> Now, I had to replicate settings in another server. First I cloned the VM of 
> the first deployment and reviewed the flentrc file, which has the following 
> lines:
>
> `[global]
> marking_names=10=0x28;12=0x30;14=0x38;16=0x40;18=0x48;20=0x50;26=0x68;34=0x88`
>
> Once everything was in place I ran the following test:
>
> `flent rrul_var -H 10.111.38.60 --test-parameter bidir_streams=5 
> --test-parameter markings=10,12,14,16,20`
>
> In principle all went fine:
>
>  
> ![Capture](https://user-images.githubusercontent.com/61751672/101208375-1ac72e00-3640-11eb-935d-8f88eb9f14b1.JPG)
>
> Simultaneously, I performed a packet capture to see the marks in the traffic 
> generated in the upload traffic as the download one doesn't seem to be marked 
> maybe due to a router blocking, but that is not an issue right now. 
>
> Looking at the upload captures I'm not seeing the outcome expected. I should 
> see values of 0x28, 0x30, 0x38, 0x40 and 0x50 in those packets. Instead, I 
> see values like 0x10, 0x0a, ox0e, etc.
>
> ![Capture](https://user-images.githubusercontent.com/61751672/101209320-8a89e880-3641-11eb-873e-ec7e87534ad5.JPG)
>
> After banging my head for a wile, I would like to ask if you see anything 
> that I may be missing in steps described above?
>
>  Any clue would be appreciated.

Hmm, seems like the configuration is correct on the Flent side. However,
note that the values you specify are the values of the entire TOS byte,
whereas what wireshark displays is only the DSCP field. So you need to
left-shift your values by two bits. I.e., if you want the DSCP field to
be 0x28, you need to set the marking to (0x28 << 2) == 0xa0...

0x28 >> 2 is 0x0a, so this seems to be consistent with what you're
seeing :)


-- 
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/215#issuecomment-739520901___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] support (re-)enabling netperf routing of UDP packets (#214)

[Toke Høiland-Jørgensen]

Jonas Smedegaard  writes:

> Output graph for an RRUL test is missing the 3 UDP ping series, but
> all other series seem to have gotten through.  Perhaps this is no
> issue in flent after all, just my cloud provider blocking UDP ping...

Or it may be that the same UDP issue you linked also affects the UDP_RR
test; I'll dig around in the netperf source code a bit and see what I
can find :)


-- 
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/214#issuecomment-739493144___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Re: [tohojo/flent] support (re-)enabling netperf routing of UDP packets (#214)

[Toke Høiland-Jørgensen]

Jonas Smedegaard  writes:

> I want to test my internet connection, and wanted to test against a
> virtual machine at Scaleway - which uses 1:1 NAT (seen from the host
> itself is a private 10.x.x.x/23 IPv4 address mapped to a public IPv4
> address) which apparently require setting [an undocumented option -R
> in
> netperf](https://stackoverflow.com/questions/11981480/error-in-running-netperf-udp-stream-over-openvpn/24211455#24211455).
>
> Please have flent support that custom "re-enable routing which we
> silently disabled for UDP) netperf option - or support passing custom
> options if you dislike hardcoding something undocumented.

According to the post you linked this is only for UDP_STREAM tests,
which we don't have any of in Flent; so why exactly is this needed?


-- 
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/214#issuecomment-738760559___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

[Flent-users] Mailing lists moved

[Toke Høiland-Jørgensen]

Hi everyone

Just a quick note to let you all know that the Flent mailing lists have
moved to a new server. I've moved over all the subscriptions, so if
you're receiving this all should be in order.

The new server is powered by mailman3, and the archives are now in
public-inbox format, available at https://lists.flent.org/.

Let me know if you have any issues!

-Toke
___
Flent-users mailing list -- flent-users@flent.org
To unsubscribe send an email to flent-users-le...@flent.org

Re: [Flent-users] [tohojo/flent] comparison with - or use of - uperf (#213)

[Toke Høiland-Jørgensen]

Jonas Smedegaard  writes:

> It seems this tool has quite some overlap with [uperf](http://www.uperf.org/).

Hmm, maybe? I'm not familiar with uperf...

> Would be nice if flent documentation mentioned how it compares to
> uperf.

... but since you seem to be, would you care to propose some text? :)


-- 
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/213#issuecomment-728367293___
Flent-users mailing list
Flent-users@flent.org
http://flent.org/mailman/listinfo/flent-users_flent.org

Re: [Flent-users] [tohojo/flent] Meaning of fields (#210)

[Toke Høiland-Jørgensen]

sebelk  writes:

> Hi,
>
> What is the meaning of fields BE, BK, CS5 and FF that it ouputs in the graphs?
> I've searched a  bit, but I didn't found anything.
> Thanks in advance!

Those are diffserv markings:
BE: Best Effort (marking 0x0)
BK: Background (AKA CS1, marking 0x20)
CS5: Class Selector 5 (marking 0xa0)
EF: Expedited Forwarding (marking 0xb8)

The particular markings were chosen so that the four flows should be put
into the four different hardware queues on Linux WiFi devices. Some ISPs
also treat diffserv-marked traffic differently, while others just clear
all markings. The RRUL test can also be used as a way to discover this...


-- 
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/210#issuecomment-646012916___
Flent-users mailing list
Flent-users@flent.org
http://flent.org/mailman/listinfo/flent-users_flent.org

Re: [Flent-users] interpreting my RRUL results

[Toke Høiland-Jørgensen]

Jeremy Marks  writes:

> Toke -
> Thank you for replying so quickly, and directly. You've been really helpful.
>
> I am a little confused about one of your answers. What I posted was
> RRUL results from directly behind my modem. But I think you are saying
> that replacing my current secondary router with something else could,
> with some tweaking, actually improve my crappy modem's performance. I
> don't understand how bloat on the modem can be addressed by the
> secondary router. Did I misunderstand?

No, you got that right. The basic thing to grasp is that bufferbloat
only occurs at the bottleneck of any given link. If your modem is not
the bottleneck, it will not queue any packets (since they can be
forwarded as fast as they come in), and you won't see any of its bloat.

So the trick is to move the bottleneck away from your modem to where we
can control the queueing. This is what the traffic shaping part of the
SQM algorithms (built-in to CAKE, or bolted-on to FQ-CoDel via the HTB
qdisc) does, and it's why you have to fiddle a bit with the bandwidth
settings: You need to shape at slightly below the rate of your actual
link to prevent the modem from being the bottleneck. And obviously you
want to get as close to the real rate to not waste any of your precious
bandwidth :)

> If I can improve bloat with a better router behind my crappy modem, I
> have been very pleased with my Ubiquiti wireless APs, and thought that
> replacing my old ASUS RT-AC66U with a beefy Edgerouter might get me
> hardware that will play nicely together.
>
> However, from reading the 4 year long thread, the process of getting
> Lochnair's port of Cake to the EdgeRouter working seems challenging
> and could be a huge time sink. Perhaps an OpenWRT router is the most
> straightforward...

Yeah, I think it may be. The hardware is not actually *that* important;
the only limitation there is that you need enough CPU to shape at the
rate of your link. Most routers should be able to handle the ~100Mbps of
total bandwidth that you need, though; IIRC that was a bit below where
we started hitting the limits of the decade-old Netgear WNDR3800... So
if you have some old hardware lying around that is supported by OpenWrt
that could be a decent place to start...

-Toke

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

Re: [Flent-users] interpreting my RRUL results

[Toke Høiland-Jørgensen]

Jeremy Marks  writes:

> This rrul test was run in a Linux Mint VM on a computer directly
> connected to my Arris BGW210 DSL modem, carrying my AT provided
> service at 75/20.
>
> My original plan was to analyze buffer bloat from machines behind my
> router that is behind this modem, in case my performance would
> increase with a Ubiquiti Edgerouter that supports shaping algorithms
> such as cake. However, I thought I had better first see what impact
> the modem itself has on the buffer bloat situation.

Good call! Always know your baseline :)

> Below is the standard unscaled triple plot from rrul. It looks pretty
> awful to me, but I am not experienced in reading these.

Yeah, looks like a pretty standard helping of bufferbloat. Notice how
the latency goes up immediately after the bulk flows start at t=5
seconds. From eyeballing, it looks like you have ~50ms of bloat on the
link. It also looks like you are not quite able to reach your 75 Mbps of
downstream bandwidth on this test, whereas the upstream rate is closer
to the nominal rate you quoted.

> I would be grateful for help in interpreting these graphs, and for any
> suggestions on what, if anything, I can do (buy an IQrouter, perhaps)?

Buying an IQrouter is the no-knobs solution. You can also see how far
you can get with the Edgerouter (not sure what that supports these
days), or you can flash OpenWrt on something and configure it manually.

> Also, I would imagine that,  the best I could do with a traffic
> shaping secondary router would be to not make things any worse. Would
> that be correct?

Nah, you should be able to get rid of most of your bloat with a suitably
configured shaper. The only question is how close to your actual link
capacity you can get away with setting the shaper to. Basically, you
set the shaper to somewhat lower than your link capacity, measure again,
and verify that it gets rid of the bloat; then keep fiddling with the
bandwidth settings (doing a binary search for the actual capacity) until
you find the maximum setting where the bloat doesn't come back. Getting
the link technology settings right helps getting you closer to the
maximum without running over.

-Toke

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

Re: [Flent-users] [tohojo/flent] Is netperf.bufferbloat.net alive? (#209)

[Toke Høiland-Jørgensen]

sebelk  writes:

> Hi, sorry for the silly question, but I've run:
>
> `flent rrul -p all_scaled -l 60 -H netperf.bufferbloat.net -t my-graph -o 
> filename.png`
>
> And get errors such as:
>
> ```
> Stdout: establish control: are you sure there is a netserver listening on 
> netperf.bufferbloat.net at port 12865?
> establish_control could not establish the control connection from 0.0.0.0 
> port 0 address family AF_INET to netperf.bufferbloat.net port 12865 address 
> family AF_INET
> ```
>
> Is 12865 the correct port?
>
> Thanks in advance!

That server sometimes runs out of bandwidth (it's hosted by a volunteer
on a bandwidth-limited connection). So if it's down, it's down; you can
try one of the others (netperf-west.bufferbloat.net or
netperf-eu.bufferbloat.net)


-- 
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/209#issuecomment-643996459___
Flent-users mailing list
Flent-users@flent.org
http://flent.org/mailman/listinfo/flent-users_flent.org

Re: [Flent-users] flent - no valid data

[Toke Høiland-Jørgensen]

Jeremy Marks  writes:

> Dear All
> I'm running flent in a Mint VM under Hyper-V in Windows 10. I am using 
> netperf.bufferbloat.net as the destination host.
> I am getting "no valid data" with rrul.
> From the log, I can see that the lrtt connection check fails, but I am not 
> knowledgeable enough to tell if anything else if failing.
> I would be very grateful for help.
> Below is the debug log. Thanks so much for your suggestions.

Seeing as both irtt and the netperf UDP flows are failing, maybe there's
simply something blocking UDP somewhere on your network?

-Toke

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

Re: [Flent-users] [tohojo/flent] Status of "--enabled-demo" mode for macOS Homebrew installs of Netperf (#207)

[Toke Høiland-Jørgensen]

Closed #207.

-- 
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/207#event-3392511353___
Flent-users mailing list
Flent-users@flent.org
http://flent.org/mailman/listinfo/flent-users_flent.org

Re: [Flent-users] [tohojo/flent] Install instructions for macOS (#208)

[Toke Høiland-Jørgensen]

Merged #208 into master.

-- 
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/pull/208#event-3392154367___
Flent-users mailing list
Flent-users@flent.org
http://flent.org/mailman/listinfo/flent-users_flent.org

Re: [Flent-users] [tohojo/flent] Install instructions for macOS (#208)

[Toke Høiland-Jørgensen]

@tohojo approved this pull request.

Great, 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/pull/208#pullrequestreview-421686330___
Flent-users mailing list
Flent-users@flent.org
http://flent.org/mailman/listinfo/flent-users_flent.org

Re: [Flent-users] [tohojo/flent] Install instructions for macOS (#208)

[Toke Høiland-Jørgensen]

@tohojo requested changes on this pull request.

Please drop those two lines, and squash everything into a single commit - other 
than that, LGTM :)

> +  
+.. code-block:: bash
+
+cd /usr/ports/net/flent && make install
+
+- **macOS:**
+
+  `Homebrew `_ and Python 3 must be installed
+  
+  You can install Python 3 easily using homebrew
+  
+.. code-block:: bash
+
+  brew install python
+  
+  A more preferred way to install Python on macOS is to use `pyenv 
`_, but this requires a bit more work to setup

I think we should just drop this line - it's more mystifying than useful...

> +  
+.. code-block:: bash
+
+cd /usr/ports/net/flent && make install
+
+- **macOS:**
+
+  `Homebrew `_ and Python 3 must be installed
+  
+  You can install Python 3 easily using homebrew
+  
+.. code-block:: bash
+
+  brew install python
+  
+  A more preferred way to install Python on macOS is to use `pyenv 
`_, but this requires a bit more work to setup

As above

-- 
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/pull/208#pullrequestreview-421395202___
Flent-users mailing list
Flent-users@flent.org
http://flent.org/mailman/listinfo/flent-users_flent.org

Re: [Flent-users] [tohojo/flent] Status of "--enabled-demo" mode for macOS Homebrew installs of Netperf (#207)

[Toke Høiland-Jørgensen]

Kris Anderson  writes:

> I want to think about the naming a bit more before I create a PR
> (undecided on `netperf-demo-mode` or if that should be something else
> like `netperf-enable-demo` or something). Do you have an opinion on
> the package name? Homebrew will use `netperf` by default, but they
> recommend using a new name to keep the installs a little simpler for
> end users.

Yeah, naming is the hardest part :)
I'm no better at it myself, so don't really have any opinion. Other than
the fact that I don't like the 'demo mode' feature name in the first
place - it has the wrong connotations, I think ("demo" sounds a bit too
much like "beta", or something unfinished).

> I also refactored the readme a bit. Not sure if you're a fan of this look, 
> but wanted to get your thoughts before I submit a PR. Here's my branch with 
> the README changes this far:
> https://github.com/kris-anderson/flent/tree/update-readme-for-macos

I think it's better to move 'other linux' up with debian/ubuntu/Arch,
but other than that, no objections :)


-- 
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/207#issuecomment-634532933___
Flent-users mailing list
Flent-users@flent.org
http://flent.org/mailman/listinfo/flent-users_flent.org

Re: [Flent-users] issues with flent setup in VM

[Toke Høiland-Jørgensen]

 writes:

> Dear All,
>
> I stood  up a guest Linux Mint VM under Hyper-V in Win 10 specifically to
> run tests with flent. However, this code:
>
>  
>
>
>
>  
>
> fails:
>
>  
>
>
>
>  
>
> I can ping netperf.buffetbloat.net just fine from the VM. I am not sure what
> lookup is failing. This is probably some basic networking setup issue with
> the VM, but it is beyond me. 
>
>  
>
> How would I go about debugging this issue?

Try lower-case 'rrul' instead of 'RRUL' :)

-Toke

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

Re: [Flent-users] [tohojo/flent] Status of "--enabled-demo" mode for macOS Homebrew installs of Netperf (#207)

[Toke Høiland-Jørgensen]

Kris Anderson  writes:

> Sounds good!
>
> I put together a working [brew
> tap](https://github.com/kris-anderson/homebrew-netperf) that should
> solve this for others. I'll take a look at creating a PR for you in
> the next couple of days to update the README and hopefully make
> running Flent easy for all macOS users using Homebrew.
>
> I tried to keep my tap true to the original formulae, so it still
> downloads the same original file from HP, but it does a find and
> replace during install to patch the file and then it adds the flag
> `--demo-mode` when compiling.

LGTM, 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/207#issuecomment-633919413___
Flent-users mailing list
Flent-users@flent.org
http://flent.org/mailman/listinfo/flent-users_flent.org