flent-users writes:
> Hi Dave,
>
>
>> On Sep 3, 2018, at 18:00, Dave Täht wrote:
>> vnv
>> Another rrul_v2 issue would be to correctly end up in all the queues on wifi.
>
> So in theory rrul_cs8 should do that... (it aims to just use one
> dscp-marked flow per class selector for a total of 8 tc
flent-users writes:
> HI Toke,
>
>
>> On Sep 3, 2018, at 19:37, Toke Høiland-Jørgensen
>> wrote:
>>
>> flent-users writes:
>>
>> > Hi Dave,
>> >
>> >
>> >> On Sep 3, 2018, at 18:00, Dave Täht wrote:
>> &
Louie Lu writes:
> Hi all,
>
> In case you want to use IRTT on an Android phone, I just make a very
> simple version of IRTT for Android, which will run `irtt server` on
> your Android phone and give you the stdout on the screen.
Cool! Does it do client mode as well? :)
-Toke
_
Louie Lu writes:
> Not now, it is in my plan to update
> , the goal is to make this app act like magiciperf*, which accept command
> line options by a input text box.
>
> *
> https://play.google.com/store/apps/details?id=com.nextdoordeveloper.miperf.miperf
Sounds great! Will look forward to taki
Dave Taht writes:
> Linode is making available nanoservers for 5/month. (was paying
> 20/month two years ago)...
>
> A) flent-singapore.taht.net. I would actually like flent.org to seize
> control of all these domains...
I can set up cnames if you send me a list :)
> B) linode's kernel ships an
Do you mean a box plot of just the bandwidth, or one that includes HTTP fetch
times?
--
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/150#issuecomment-430749681__
Closed #149 via 79f013ce5742682f65f47bf957a286759628b61a.
--
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/149#event-1910451575___
Flent-users
I closed it because I added the test... :)
--
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/149#issuecomment-430910772___
Flent-users mailing
Yeah, the box_combine plot is not terribly useful with a single test. It plots
the distribution of averages over several test runs; so try it with multiple
files.
I'll add the other type as well :)
--
You are receiving this because you are subscribed to this thread.
Reply to this email direct
eriktar writes:
> Perhaps I'm doing it wrong, but I still find box more readable.
Right, sure. Be aware that because box plots express percentiles, they
can be misleading for data sets that do not consist of independent
samples, such as timeseries data points from the same download.
I'm not say
Ah, awesome, thanks!
Care to submit a patch to the documentation (doc/_templates/index.html
and doc/quickstart.rst) with a note for FreeBSD users? :)
--
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
Great, thanks!
One question: Does this work for any version of FreeBSD, or just the latest?
--
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/153#issuecomment-432558706
Things broke when matplotlib 2 was released, so wouldn't be surprised that the
same thing happened for v3. I'll look into 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/154#issu
Ryan Steinmetz writes:
> Should work on any currently supported release. Today, this would be:
> Production: 11.2, 10.4.
>
> https://www.freebsd.org/ has the information
Cool, thanks!
--
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it
Closed #153 via c928c03a301258c26c7d045c74ecce6dfeaa3d5a.
--
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/153#event-1923148303___
Flent-users ma
(accidentally dropped the list, adding it back)
Dave Taht writes:
> On Sat, Oct 27, 2018 at 5:33 AM Toke Høiland-Jørgensen wrote:
>>
>> Dave Taht writes:
>>
>> > so I just upgraded to ubuntu 18.04 finally and tried to get flent running.
>> >
>>
Closed #154 via 119ac78345ea032c6fcbd1caadbb05fdc85da59b.
--
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/154#event-1951345846___
Flent-users
Leslie Monis writes:
> Hi,
>
> I have encountered an issue when running the tcp_nup tests on newer
> versions of Linux, specifically v4.16 and up. It seems that flent
> isn't able to get enough data points (for upload throughput) for
> individual TCP flows. This is causing the median TCP upload s
Hmm, I think this could be caused by this commit:
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=0a6b2a1dc2a2105f178255fe495eb914b09cb37a
...which makes TCP always use GSO.
This was introduced in Linux 4.17; is there any chance you could test that the
difference ap
Can you share the data file you used to generate the 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/158#issuecomment-452265859___
Flen
So, finally got around to look at this. Wow, that's an annoying change for the
Flent use case. Did you find a workaround?
--
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/146#issuecommen
flent-users writes:
> On Tue, Jan 8, 2019 at 3:34 AM Toke Høiland-Jørgensen <
> notificati...@github.com> wrote:
>
>> So, finally got around to look at this. Wow, that's an annoying change for
>> the Flent use case. Did you find a workaround?
>>
>
> O
Leslie Monis writes:
>> Hmm, I think this could be caused by this commit:
> https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=0a6b2a1dc2a2105f178255fe495eb914b09cb37a
>>
>>...which makes TCP always use GSO.
>
> Yes, the commit does seem to cause the issue.
Great, than
Pete Heist writes:
> There doesn’t seem to be an irtt server running on flent-london. Is it
> off, or did it go down and was there anything logged?
No idea. Dave?
-Toke
___
Flent-users mailing list
Flent-users@flent.org
http://flent.org/mailman/listi
cklatsky writes:
> Hi,
>
> I have a new setup that I am trying to get Flent gui to run, so I may view
> test output files. The setup is Cygwin environment running on Windows 10.
> Cygwin definition copied below from Wikipedia. I have all the packages
> installed successfully, but I get this
Sebastian Moeller writes:
> Dear all,
>
> Loganaden remined me of a pet project of mine, deducing the
> per-packet-overhead from non ATM/AAL5 carriers. Now, I believe I have
> an approach that might be worth exploring*, but that requires me to
> saturate my link with packets of different packet s
cklatsky writes:
> Hi,
>
> Do any of the supplied tests utilize netperf generated latency
> statistics (-j option) for the latency value?
Not currently, no. Not sure how Netperf actually gets those values for
TCP either. With the --socket-stats option you can get the values
reported by the kerne
Sebastian Moeller writes:
> So I try to run:
>
> bash-3.2$ ./run-flent --ipv4 -l 60 -H netperf-eu.bufferbloat.net
> --test-parameter=download_streams=16 tcp_ndown
> --remote-metadata=root@192.168.1.1 -D . -t
> IPv4_SQM_cake_layer-cake_LLA-ETH_OH34_U098pct9500of40309K-D090pct49000of109924K_hms
acdcjavier writes:
> Hello
> Is it possible to force Iperf testing instead netperf?
>
> I need to make some thoughtput test in my LAN configuration, but
> netperf is older and Iperf is more compatible with Linux/windows.
Not generally, no. Netperf has quite a few options that iperf lacks, and
it
rayhammerco writes:
> Hello, Im trying to run Flent on MACos and am seeing the following
> errors. Im hoping someone can point me in the right direction on this.
> Thank you.
Hmm, I think this is fixed already; could you try the latest git version
and see if the problem persists, please?
--
Y
Ah, no it isn't! Seems there's a new bug with matplotlib 3.1... Will
investigate!
--
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/162#issuecomment-494750344_
Closed #162 via 6f89a1b5cf4766507dd7d96e8c9b58ab2cca51ae.
--
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/162#event-235077___
Flent-users
No, this is just configuration for the Travis CI tests; no point in running the
tests for matplotlib 3.1 on Python 3.5 as that is not supported anyway...
When running Flent yourself, you just need matplotlib installed in whichever
Python version you're using...
--
You are receiving this becaus
Pete Heist writes:
> I have ss installed and am using --socket-stats, but for some reason
> tcp_rtt plots don't seem to be working (using rrul_be but I need it
> ultimately for tcp_1up_noping). I'm using flent from HEAD and python
> 3.7.3 on Ubuntu 19.04. Do the attached .flent.gz files and outpu
Pete Heist writes:
> Ah, that is very likely that CONFIG_INET_DIAG needs to be set, so I'll
> re-compile my kernel with that...
Do let me know if that works; I don't get RTTs on my Arch kernel either,
it would seem (though I don't have the permission issue).
Will try to figure out if anything c
Luis Glass writes:
> Hi - I'm a new user and still learning flent. I'm trying to use the
> batchfile.example using the command "flent -B batchfile.example", and am
> getting the error "ERROR: Missing test name".
>
> Am I missing something in the command syntax? It seems like the test name
> is d
Closed #164.
--
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/164#event-2446844848___
Flent-users mailing list
Flent-users@flent.org
http://flent
Thanks for the report, and fix!
There were a few style issues with the pull request (missing DCO, no commit
message, you forgot to move the metadata saving of the command), so I just
committed this fix instead of merging the pull request.
--
You are receiving this because you are subscribed to t
Fixed in 276abd7e28d264e76d8b739f7e3867710cfbedbc
--
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/164#issuecomment-506679032___
Flent-users mail
Pete Heist writes:
> I'm building flent from head with make and installing with make install. For
> some reason when I start `flent-gui` with or without arguments (here started
> with --debug) I get:
>
> ```
> % flent-gui --debug
> Started Flent 1.2.2-git-a8d31e9 using Python 3.7.3.
> Initialis
Closed #165 via c732080ebd9e0b7faafbc24f5d1a125c18ce40ef.
--
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/165#event-2461204162___
Flent-users
Closed #166 via 48ecf31d73a27f97cd69a18bf614a6ff331233d4.
--
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/166#event-2467392422___
Flent-users
Oops, this is a fallout from the conversion of that output to the logic that
allows --data-cutoff. Will fix, 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/166#issuecomment-5
Released on 2019-07-09.
Note that this will be the last release with Python 2 support!
Changes since v1.2.2 include:
- New --data-cutoff parameter filter data outside a specified time interval when
plotting or formatting. This is useful for, e.g., showing flow behaviour after
steady state ha
Hi everyone
This is to announce the release of Flent v1.3.0. The automated Github
announcement already went out, but I figured I'd do a manual one as
well.
Note that this will be the last release of Flent to retain support for
Python 2. The upstream end-of-life for Python 2 is now only around 6
m
Closed #65.
--
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/65#event-2470365462___
Flent-users mailing list
Flent-users@flent.org
http://flent
We actually support this with the 'markings' test parameter now :)
--
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/65#issuecomment-509765557__
Pete Heist writes:
> I compiled on some new test boxes from head (83d14d68) and ran a test. flent
> -i for the text output worked but when I use flent -gui I see the following
> errors before it freezes:
>
> ```
> ERROR: Error while loading data file: '
> returned a result with an error set'.
Pete Heist writes:
> Hmm, that made me realize it's working in Linux but not on my Mac, with
> slightly different versions of Python and PyQt5:
>
> Not working (on MacOS 10.13.6):
> ```
> % flent --version
> Flent v1.9.9-git-83d14d6.
> Running on Python 3.7.4 (default, Jul 9 2019, 18:15:00) [C
Pete Heist writes:
> Against better judgment I upgraded PyQt5 to 5.13.0 on both the Mac and
> Linux and still see the problem on Mac but not on Linux, so that
> suggests it's not directly related to the PyQt 5 version.
Does it happen if you only load a single data file?
--
You are receiving t
Pete Heist writes:
> Ok, the problem doesn't happen with PySide2. It's likely with recent
> versions of PyQt5, although I'm not exactly sure of the first version
> where it started. It's also possible the bug doesn't happen on MacOS
> 10.14 (Mojave), but my 2011 hardware is pinned to 10.13 so I c
Pete Heist writes:
> Ah, ok, well it looks like PySide2 has a more permissive license
> anyway, although I guess that shouldn't affect flent. I haven't
> noticed any other problems using it, although my testing has hardly
> been extensive.
Yeah, PySide2 are the officially supported Python bindin
Sebastian Moeller writes:
> Okay, same issue on macosx 10.14.5. Now with the latest changes and PySide2
> installed I get:
>
>
> ./run-flent --version
> Flent v1.9.9-git-43e16d5.
> Running on Python 3.7.4 (default, Jul 11 2019, 01:08:00) [Clang 10.0.1
> (clang-1001.0.46.4)].
> Using matplotlib
Sebastian Moeller writes:
> Mmmh, so I installed qtpy (and dependencies) I now get:
>
> ./run-flent --gui
> Started Flent 1.9.9-git-43e16d5 using Python 3.7.4.
> Initialised matplotlib v3.1.1 on numpy v1.16.4.
> GUI loaded. Using Qt through pyqt5 v5.12.3.
> objc[13628]: Class FIFinderSyncExtensio
Pete Heist writes:
> It looks like it was still using PyQt5 in your case, so make sure to first do:
>
> ```
> export QT_API= pyside2
> ```
>
> If this is a long-term fix for Mac, maybe that var could be set on
> MacOS when sys.platform is darwin, assuming we have a chance to do so
> before those
Hi
I'm the maintainer of the Flent tool, which uses PyQt for its GUI. I
recently got a bug report which showed odd behaviour on OS X.
Specifically, setupUi() returns:
SystemError: returned a result with an
error set
With what appears to be an underlying: AttributeError: __getattribute__
The e
Hi
I'm the maintainer of the Flent tool, which uses PyQt for its GUI. I
recently got a bug report which showed odd behaviour on OS X.
Specifically, setupUi() returns:
SystemError: returned a result with an
error set
With what appears to be an underlying: AttributeError: __getattribute__
The e
Merged #168 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/168#event-2497729311___
Flent-users mailing list
Flent-users@flent.org
Heh, division, multiplication, same same, right? Oops. Thanks for catching (and
fixing) 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/pull/168#issuecomment-513497022__
Pete Heist writes:
> Is there a way I could start two latency measurement flows, either
> both ICMP or both UDP (ideally it should be the same), each of which
> has a different value for the tos field, alongside the tcp_nup test?
See 878f297cd336539189fe577ea82c1037ed136b20 which adds the
ping_m
Pete Heist writes:
> Thanks! That's working over the wire, and I've managed to add plots
> for throughput+pings and delivery+pings, so this is working.
Awesome!
> In general, is it worth upstreaming such custom plots? If I don't, I
> sometimes forget and blow them away when cloning a fresh tree
Dave Täht writes:
> It sounds like a rrul_dualq test is now feasible? The rrul is notorious for
> blowing up
> single queue designs, which take forever to gain control of the queue and
> drop a zillion
> ack packets to do so. Long term this test would also be useful in
> evaluating accecn.
Wha
dignacioconde writes:
> root@n1:/tmp/pycore.50542/n1.conf# flent -H 10.0.0.21 udp_flood --debug-error
> Started Flent 1.3.0 using Python 3.6.8.
> Starting udp_flood test. Expected run time: 70 seconds.
> WARNING: Found iperf binary (), but it does not have an --enhancedreports
> option. Not usin
Closed #172.
--
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/172#event-2507133740___
Flent-users mailing list
Flent-users@flent.org
http://flent
Thanks! Merged in slightly modified form as
d33fd063f6f32ddaff4aeb314504d251c8d8a480
--
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/172#issuecomment-514732426
Pete Heist writes:
> In tcp_nup.conf, I'm trying to create a dual axis plot that combines delivery
> rate on one axis along with ping and TCP RTT on the second axis, attempting
> like this:
>
> ```
> ('tcp_delivery_with_ping_and_tcp_rtt',
> {'description': 'TCP delivery rate with ping
Closed #173 via 1ff2d17f111d46c39e851d96c19ccdd97f931045.
--
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/173#event-2509106170___
Flent-users
Closed #174.
--
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/174#event-2509847843___
Flent-users mailing list
Flent-users@flent.org
http://flent
Merged as c8d9f8c3aaf526eb3d0c96ebe27be53deb305a85
--
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/174#issuecomment-515083193___
Flent-users mai
Bicobus writes:
> Isn't there a bugtracker for PyQt5 somewhere or is the only way to
> reach riverbankcomputing to post on their mailing list?
Couldn't find one. Which is one of the reasons why I figure it's a good
idea to focus on pyside2 instead :)
--
You are receiving this because you comm
Justin LeFebvre writes:
> I'm seeing the same there here. An interesting thing to note is that
> this seems to be an issue with PyQt5 and python 3.7.4. When running
> with python 3.7.3, I'm not seeing the issue.
Ah, that's interesting! Still not something that's likely we can work
around from Fl
Certainly don't mind taking this, no! Only concern is that the "later is
preferred" resolution order is the opposite of multiple inheritance in Python,
which may be confusing? On the other hand, it's not really code, and I can see
why it makes sense this way as well.
Was there any particular re
Right, so the code already does new['inherits'] = parent, child; this
corresponds to your ordering, so let's just go with 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/pull/175#issuec
Merged as 835d1d9917f32cb2162710133a0cc2292d308860
--
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/175#issuecomment-526200503___
Flent-users mai
Closed #175.
--
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/175#event-2594119495___
Flent-users mailing list
Flent-users@flent.org
http://flent
Merged #176 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/176#event-2596721782___
Flent-users mailing list
Flent-users@flent.org
Rich Brown writes:
> Spurred on by Dave's [Coffee Shop Bloat
> Test](https://lists.bufferbloat.net/pipermail/bloat/2019-September/009336.html)
> script, I decided to try it on my Mac (10.12.6) with Flent 1.3.0. This
> command seemed to work fine:
>
> `flent -x -H flent-fremont.bufferbloat.net
Rich Brown writes:
> Hmmm, not this time... What else could I check?
>
> ![image](https://user-images.githubusercontent.com/1094930/64439228-7e5a5880-d097-11e9-9068-1f80fd634be7.png)
If you turn on debug logging during the test run (-L option), that log
will include the actual output of the ping
> Would you be ok if I do that, or do you have any other suggestions for
> this?
Not opposed to having an option; --batch-no-timestamp ?
--
You are receiving this because you commented.
Reply to this email directly or view it on GitHub:
https://github.com/tohojo/flent/issues/179#issuecomment-5
Rich Brown writes:
> Here's the log file: https://pastebin.com/aBXCgjD1
>
> NB I'll be away for the weekend, so won't follow up for a while.
> Thanks.
Looks like those are the actual ping times reported by fping:
[1567795827.309700] flent-fremont.bufferbloat.net : [175], 84 bytes, 95.5 ms
(104
Merged #180 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/180#event-2616232665___
Flent-users mailing list
Flent-users@flent.org
Rich Brown writes:
> Ahah! I see the difference. I was testing against `gstatic.com` with
> Terminal; Flent was testing against `flent-fremont.bufferbloat.net`
>
> This seems a pretty big disparity between test results. Is there an
> option to tell Flent to ping a more local host? Thanks.
You c
Rich Brown writes:
> Hmmm... The `--test-parameter=ping_hosts=gstatic.com` information
> appears in the command-line summary but not in the charts. Is this
> correct?
>
> If so, the charts display significantly different latency depending on
> which -H host gets used. The Terminal display (bottom
Rich Brown writes:
>> > 2. Why is there a discrepency between the
> `--test-parameter=ping_hosts=...` reading (showing 30-38 msec in my
> tests) and a separate, concurrent fping test (see original post, where
> it was running 20-22 msec)?
>
>> Good question. IPv4 vs IPv6 maybe?
>
> I don't thi
Rich Brown writes:
>> > 1. Could flent charts use the `--test-parameter=ping_hosts=...` instead of
>> > values from the -H host?
>
>> Well, it should probably show both by default. For now, there's a separate
>> plot type you can use to see the extra flows (-p ping_extra).
>
> I don't see how t
Rich Brown writes:
> **Update:** I chose **View -> Plot selector** and got the plot selector to
> appear as a free-standing window. I clicked in its title bar and dragged it
> to the main window, and they merged to show the image below.
>
> New question: Is it possible to select two plots (say
Rich Brown writes:
>> Doing this dynamically in the GUI would be a nice feature, though, but
> not quite trivial unfortunately...
>
> Yes. I once worked on the plotting portion of a program, and nothing is
> trivial.
>
> So I come back to my request to select the ping_host value in lieu of late
Rich Brown writes:
> Update: I'm trying to get repeatable test showing a difference of latency
> between the Flent ping_host values and a separate/concurrent fping run. (I
> saw it two days ago, but can't reproduce now... This is roundabout way to say
> that, for the moment, the Flent ping_hos
Rich Brown writes:
> Actually, I want the ability to plot the "best"`ping_host` (that is, lowest
> latency host), since that's going to display the biggest variance. A few
> milliseconds of additional bloat added to the 90msec from my house to Fremont
> is easy to ignore. Tacking a few millise
Rich Brown writes:
> Update: dtaht (in
> https://github.com/tohojo/flent/issues/178#issuecomment-528885331)
> said of the ping times, "I have seen this statistic be wildly wrong in
> the summary but correct in the graph."
Yes, this can be the case because of the way the summary values are
comput
Closed #182 via ff465203dd7de78d87e0d88717068621b8159c28.
--
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/182#event-2632860804___
Flent-users
Rich Brown writes:
>> One way to achieve something like it is to add another test parameter
> that would override the ping destination of the existing tests. I.e.,
> instead of adding a new ping host, you'd just get the ability to
> override the one(s) already defined by the test...
>>
>> Would
Rich Brown writes:
> I don't pretend to understand the metadata - there is a lot stuff in
> the JSON file - but I don't think including the information in the
> filename would be necessary.
>
> I don't foresee a transition problem here. Flent can continue to
> default to using the -H host for its
Rich Brown writes:
> Sorry to re-open, but the new version (Flent 1.9.9-git-ff46520) seems
> to fix the upload values for the Average - the individual streams sum
> to the displayed sum.
>
> But the Median values still don't look right. Here are two runs of the
> same command:
Ah, no, the median
That should all be fixable, at least for new tests. The summary output just
prints data series in the order they are specified in the data file, so harder
to re-sort old files...
On 15 September 2019 13:51:48 CEST, Rich Brown wrote:
>Using Flent 1.9.9-git-2a08dd9:
>
>I have been looking hard at
Hmm, hard to say for sure without data and log files, but this might be an
artifact of the netperf UDP RR test. I would expect you'd get better data using
irtt...
On 15 September 2019 13:56:31 CEST, Rich Brown wrote:
>Flent 1.9.9-git-2a08dd9:
>
>I am confused about the following. Might they be
Ah bugger, seems I was wrong and that the series are actually just being sorted
alphabetically. So guess I'll have to figure out a way to define a sorting order
that makes more sense...
--
You are receiving this because you commented.
Reply to this email directly or view it on GitHub:
https://git
Rich Brown writes:
> I'm all for gnarly hacks... Is it possible to change the names to "TCP
> upload::avg" and "TCP upload::sum"? Or change the individual tests to
> "TCP upload-N" and the "TCP upload-avg" and "TCP upload-sum"?
Nah, that won't work. For one thing, it would break compatibility wi
Rich Brown writes:
>> Rich said:
>> ... You can also tell me to get a life :-)
>
> Really - I can live without this. This is the kind of request that
> might possibly be cool (perhaps), but is hardly critical. Given that
> it could take a bunch of time to implement, I would counsel that you
> wa
Rich Brown writes:
> OK New run, this time with result and log files... (Charts are
> substantially the same as above.) Thanks!
Aha! That's useful!
Right, so there are two things going on here:
- The MEAN_VALUE stored as metadata has not been transformed from
transactions/second to ms.
- Th
201 - 300 of 490 matches
Mail list logo