[Bloat] Netperf re-licensed as MIT

[Bloat] Netperf re-licensed as MIT

[Toke Høiland-Jørgensen]

Hopefully this means we can get it packaged for the distros that have
thus far refused to because of the license - i.e., Debian and Fedora!

-Toke

Re: [Bloat] Netperf re-licensed as MIT

[Klatsky, Carl]

Toke,

How you see this coming into Flent?  My understanding is that for latency & load test like RRUL:

-iperf provides the bandwidth portion
-IRTT provides the latency portion, and if IRTT is not found, the test falls back to ICMP for the latency check

Would netperf replace iperf for the bandwidth portion?  Do you see it being used for the latency portion?

Thanks,
Carl

-----Original Message-----
From: Bloat <bloat-bounces@lists.bufferbloat.net> On Behalf Of Toke Høiland-Jørgensen via Bloat
Sent: Friday, March 26, 2021 7:32 PM
To: bloat <bloat@lists.bufferbloat.net>
Subject: [Bloat] Netperf re-licensed as MIT

Hopefully this means we can get it packaged for the distros that have thus far refused to because of the license - i.e., Debian and Fedora!

-Toke
_______________________________________________
Bloat mailing list
Bloat@lists.bufferbloat.net
https://urldefense.com/v3/__https://lists.bufferbloat.net/listinfo/bloat__;!!CQl3mcHX2A!W_yWwnGSD6VuZvpn0BPv7Ta9GBP3f2dRWeJHPLS7RrrpBZ5gpyj5DZGnRYIfWVR5TXjR$

Re: [Bloat] Netperf re-licensed as MIT

[Dave Taht]

so glad to hear the license has been fixed.

carl, iperf is only used in a few of flent's tests. We trusted netperf
- as did the linux kernel developers - a lot further than all the
iperf variants combined - at the time we started work on flent.

I would not mind us somehow developing a drop in replacement for
netperf, perhaps leveraging the irtt codebase, as I have a long
standing desire to be able to reliably measure latencies and the
output of TCP_INFO in "some" tool below 250us.

iperf has come a long way, but getting to where I could trust the
largely "academic" codebase it was would take a ton of benchmarking.

On Sat, Mar 27, 2021 at 2:09 PM Klatsky, Carl via Bloat
<bloat@lists.bufferbloat.net> wrote:
>
> Toke,
>
> How you see this coming into Flent?  My understanding is that for latency & load test like RRUL:
>
> -iperf provides the bandwidth portion
> -IRTT provides the latency portion, and if IRTT is not found, the test falls back to ICMP for the latency check
>
> Would netperf replace iperf for the bandwidth portion?  Do you see it being used for the latency portion?
>
> Thanks,
> Carl
>
> -----Original Message-----
> From: Bloat <bloat-bounces@lists.bufferbloat.net> On Behalf Of Toke Høiland-Jørgensen via Bloat
> Sent: Friday, March 26, 2021 7:32 PM
> To: bloat <bloat@lists.bufferbloat.net>
> Subject: [Bloat] Netperf re-licensed as MIT
>
> Hopefully this means we can get it packaged for the distros that have thus far refused to because of the license - i.e., Debian and Fedora!
>
> -Toke
> _______________________________________________
> Bloat mailing list
> Bloat@lists.bufferbloat.net
> https://urldefense.com/v3/__https://lists.bufferbloat.net/listinfo/bloat__;!!CQl3mcHX2A!W_yWwnGSD6VuZvpn0BPv7Ta9GBP3f2dRWeJHPLS7RrrpBZ5gpyj5DZGnRYIfWVR5TXjR$
> _______________________________________________
> Bloat mailing list
> Bloat@lists.bufferbloat.net
> https://lists.bufferbloat.net/listinfo/bloat



-- 
"For a successful technology, reality must take precedence over public
relations, for Mother Nature cannot be fooled" - Richard Feynman

dave@taht.net <Dave Täht> CTO, TekLibre, LLC Tel: 1-831-435-0729

Re: [Bloat] Netperf re-licensed as MIT

[Dave Taht]

most of the latency related portions of flent use irtt, which I think
is a lot more solid than anything else.

On Sat, Mar 27, 2021 at 2:44 PM Dave Taht <dave.taht@gmail.com> wrote:
>
> so glad to hear the license has been fixed.
>
> carl, iperf is only used in a few of flent's tests. We trusted netperf
> - as did the linux kernel developers - a lot further than all the
> iperf variants combined - at the time we started work on flent.
>
> I would not mind us somehow developing a drop in replacement for
> netperf, perhaps leveraging the irtt codebase, as I have a long
> standing desire to be able to reliably measure latencies and the
> output of TCP_INFO in "some" tool below 250us.
>
> iperf has come a long way, but getting to where I could trust the
> largely "academic" codebase it was would take a ton of benchmarking.
>
> On Sat, Mar 27, 2021 at 2:09 PM Klatsky, Carl via Bloat
> <bloat@lists.bufferbloat.net> wrote:
> >
> > Toke,
> >
> > How you see this coming into Flent?  My understanding is that for latency & load test like RRUL:
> >
> > -iperf provides the bandwidth portion
> > -IRTT provides the latency portion, and if IRTT is not found, the test falls back to ICMP for the latency check
> >
> > Would netperf replace iperf for the bandwidth portion?  Do you see it being used for the latency portion?
> >
> > Thanks,
> > Carl
> >
> > -----Original Message-----
> > From: Bloat <bloat-bounces@lists.bufferbloat.net> On Behalf Of Toke Høiland-Jørgensen via Bloat
> > Sent: Friday, March 26, 2021 7:32 PM
> > To: bloat <bloat@lists.bufferbloat.net>
> > Subject: [Bloat] Netperf re-licensed as MIT
> >
> > Hopefully this means we can get it packaged for the distros that have thus far refused to because of the license - i.e., Debian and Fedora!
> >
> > -Toke
> > _______________________________________________
> > Bloat mailing list
> > Bloat@lists.bufferbloat.net
> > https://urldefense.com/v3/__https://lists.bufferbloat.net/listinfo/bloat__;!!CQl3mcHX2A!W_yWwnGSD6VuZvpn0BPv7Ta9GBP3f2dRWeJHPLS7RrrpBZ5gpyj5DZGnRYIfWVR5TXjR$
> > _______________________________________________
> > Bloat mailing list
> > Bloat@lists.bufferbloat.net
> > https://lists.bufferbloat.net/listinfo/bloat
>
>
>
> --
> "For a successful technology, reality must take precedence over public
> relations, for Mother Nature cannot be fooled" - Richard Feynman
>
> dave@taht.net <Dave Täht> CTO, TekLibre, LLC Tel: 1-831-435-0729



-- 
"For a successful technology, reality must take precedence over public
relations, for Mother Nature cannot be fooled" - Richard Feynman

dave@taht.net <Dave Täht> CTO, TekLibre, LLC Tel: 1-831-435-0729

Re: [Bloat] Netperf re-licensed as MIT

[Toke Høiland-Jørgensen]

Dave Taht <dave.taht@gmail.com> writes:

> so glad to hear the license has been fixed.
>
> carl, iperf is only used in a few of flent's tests. We trusted netperf
> - as did the linux kernel developers - a lot further than all the
> iperf variants combined - at the time we started work on flent.

And, more importantly, netperf provides a lot of features that iperf
doesn't; such as dumping TCP info, setting congestion control, etc -
full list here:

https://lists.bufferbloat.net/pipermail/make-wifi-fast/2020-January/002648.html

-Toke

Re: [Bloat] Netperf re-licensed as MIT

[Aaron Wood]

[-- Attachment #1: Type: text/plain, Size: 2940 bytes --]

iperf3 isn’t “academic”, but is more focused on scientific computing (ESNet
pushes a LOT of data CERN around, on 100Gbps backbones).

But that also skews their usage/needs.  Very high throughput bulk transfers
with long durations, over mixed systems. Not as many concerns about
latency, except in that latency can cause messes with congestion control.

On Sat, Mar 27, 2021 at 2:45 PM Dave Taht <dave.taht@gmail.com> wrote:

> so glad to hear the license has been fixed.
>
> carl, iperf is only used in a few of flent's tests. We trusted netperf
> - as did the linux kernel developers - a lot further than all the
> iperf variants combined - at the time we started work on flent.
>
> I would not mind us somehow developing a drop in replacement for
> netperf, perhaps leveraging the irtt codebase, as I have a long
> standing desire to be able to reliably measure latencies and the
> output of TCP_INFO in "some" tool below 250us.
>
> iperf has come a long way, but getting to where I could trust the
> largely "academic" codebase it was would take a ton of benchmarking.
>
> On Sat, Mar 27, 2021 at 2:09 PM Klatsky, Carl via Bloat
> <bloat@lists.bufferbloat.net> wrote:
> >
> > Toke,
> >
> > How you see this coming into Flent?  My understanding is that for
> latency & load test like RRUL:
> >
> > -iperf provides the bandwidth portion
> > -IRTT provides the latency portion, and if IRTT is not found, the test
> falls back to ICMP for the latency check
> >
> > Would netperf replace iperf for the bandwidth portion?  Do you see it
> being used for the latency portion?
> >
> > Thanks,
> > Carl
> >
> > -----Original Message-----
> > From: Bloat <bloat-bounces@lists.bufferbloat.net> On Behalf Of Toke
> Høiland-Jørgensen via Bloat
> > Sent: Friday, March 26, 2021 7:32 PM
> > To: bloat <bloat@lists.bufferbloat.net>
> > Subject: [Bloat] Netperf re-licensed as MIT
> >
> > Hopefully this means we can get it packaged for the distros that have
> thus far refused to because of the license - i.e., Debian and Fedora!
> >
> > -Toke
> > _______________________________________________
> > Bloat mailing list
> > Bloat@lists.bufferbloat.net
> >
> https://urldefense.com/v3/__https://lists.bufferbloat.net/listinfo/bloat__;!!CQl3mcHX2A!W_yWwnGSD6VuZvpn0BPv7Ta9GBP3f2dRWeJHPLS7RrrpBZ5gpyj5DZGnRYIfWVR5TXjR$
> > _______________________________________________
> > Bloat mailing list
> > Bloat@lists.bufferbloat.net
> > https://lists.bufferbloat.net/listinfo/bloat
>
>
>
> --
> "For a successful technology, reality must take precedence over public
> relations, for Mother Nature cannot be fooled" - Richard Feynman
>
> dave@taht.net <Dave Täht> CTO, TekLibre, LLC Tel: 1-831-435-0729
> _______________________________________________
> Bloat mailing list
> Bloat@lists.bufferbloat.net
> https://lists.bufferbloat.net/listinfo/bloat
>
-- 
- Sent from my iPhone.

[-- Attachment #2: Type: text/html, Size: 4507 bytes --]

Re: [Bloat] Netperf re-licensed as MIT

[Aaron Wood]

[-- Attachment #1: Type: text/plain, Size: 3598 bytes --]

One of my long concerns with the RRUL test is that the ICMP ping test
portion is not isochronous, and runs at a variable rate based on rtt, which
means that it uses more/less bandwidth as an inverse function of rtt, and
that makes it harder to compare the actual goodput of the tcp streams
running in parallel.  (When the top-line total bw goes down when latency
goes down, because the icmp and udp ping tests are using more bandwidth
that isn’t accounted for in the bw totals)

On Sat, Mar 27, 2021 at 2:45 PM Dave Taht <dave.taht@gmail.com> wrote:

> most of the latency related portions of flent use irtt, which I think
> is a lot more solid than anything else.
>
> On Sat, Mar 27, 2021 at 2:44 PM Dave Taht <dave.taht@gmail.com> wrote:
> >
> > so glad to hear the license has been fixed.
> >
> > carl, iperf is only used in a few of flent's tests. We trusted netperf
> > - as did the linux kernel developers - a lot further than all the
> > iperf variants combined - at the time we started work on flent.
> >
> > I would not mind us somehow developing a drop in replacement for
> > netperf, perhaps leveraging the irtt codebase, as I have a long
> > standing desire to be able to reliably measure latencies and the
> > output of TCP_INFO in "some" tool below 250us.
> >
> > iperf has come a long way, but getting to where I could trust the
> > largely "academic" codebase it was would take a ton of benchmarking.
> >
> > On Sat, Mar 27, 2021 at 2:09 PM Klatsky, Carl via Bloat
> > <bloat@lists.bufferbloat.net> wrote:
> > >
> > > Toke,
> > >
> > > How you see this coming into Flent?  My understanding is that for
> latency & load test like RRUL:
> > >
> > > -iperf provides the bandwidth portion
> > > -IRTT provides the latency portion, and if IRTT is not found, the test
> falls back to ICMP for the latency check
> > >
> > > Would netperf replace iperf for the bandwidth portion?  Do you see it
> being used for the latency portion?
> > >
> > > Thanks,
> > > Carl
> > >
> > > -----Original Message-----
> > > From: Bloat <bloat-bounces@lists.bufferbloat.net> On Behalf Of Toke
> Høiland-Jørgensen via Bloat
> > > Sent: Friday, March 26, 2021 7:32 PM
> > > To: bloat <bloat@lists.bufferbloat.net>
> > > Subject: [Bloat] Netperf re-licensed as MIT
> > >
> > > Hopefully this means we can get it packaged for the distros that have
> thus far refused to because of the license - i.e., Debian and Fedora!
> > >
> > > -Toke
> > > _______________________________________________
> > > Bloat mailing list
> > > Bloat@lists.bufferbloat.net
> > >
> https://urldefense.com/v3/__https://lists.bufferbloat.net/listinfo/bloat__;!!CQl3mcHX2A!W_yWwnGSD6VuZvpn0BPv7Ta9GBP3f2dRWeJHPLS7RrrpBZ5gpyj5DZGnRYIfWVR5TXjR$
> > > _______________________________________________
> > > Bloat mailing list
> > > Bloat@lists.bufferbloat.net
> > > https://lists.bufferbloat.net/listinfo/bloat
> >
> >
> >
> > --
> > "For a successful technology, reality must take precedence over public
> > relations, for Mother Nature cannot be fooled" - Richard Feynman
> >
> > dave@taht.net <Dave Täht> CTO, TekLibre, LLC Tel: 1-831-435-0729
>
>
>
> --
> "For a successful technology, reality must take precedence over public
> relations, for Mother Nature cannot be fooled" - Richard Feynman
>
> dave@taht.net <Dave Täht> CTO, TekLibre, LLC Tel: 1-831-435-0729
> _______________________________________________
> Bloat mailing list
> Bloat@lists.bufferbloat.net
> https://lists.bufferbloat.net/listinfo/bloat
>
-- 
- Sent from my iPhone.

[-- Attachment #2: Type: text/html, Size: 5353 bytes --]

Re: [Bloat] Netperf re-licensed as MIT

[Dave Taht]

irtt is used if available.

I agree that the ping method is poor at lower scales. The ping volume
also accounts for more bandwidth the lower the rtt which rrul does not
measure. I would not mind trying to produce a rrul2021 test that
updated it better for modern conditions.

On Mon, Mar 29, 2021 at 11:29 AM Aaron Wood <woody77@gmail.com> wrote:
>
> One of my long concerns with the RRUL test is that the ICMP ping test portion is not isochronous, and runs at a variable rate based on rtt, which means that it uses more/less bandwidth as an inverse function of rtt, and that makes it harder to compare the actual goodput of the tcp streams running in parallel.  (When the top-line total bw goes down when latency goes down, because the icmp and udp ping tests are using more bandwidth that isn’t accounted for in the bw totals)
>
> On Sat, Mar 27, 2021 at 2:45 PM Dave Taht <dave.taht@gmail.com> wrote:
>>
>> most of the latency related portions of flent use irtt, which I think
>> is a lot more solid than anything else.
>>
>> On Sat, Mar 27, 2021 at 2:44 PM Dave Taht <dave.taht@gmail.com> wrote:
>> >
>> > so glad to hear the license has been fixed.
>> >
>> > carl, iperf is only used in a few of flent's tests. We trusted netperf
>> > - as did the linux kernel developers - a lot further than all the
>> > iperf variants combined - at the time we started work on flent.
>> >
>> > I would not mind us somehow developing a drop in replacement for
>> > netperf, perhaps leveraging the irtt codebase, as I have a long
>> > standing desire to be able to reliably measure latencies and the
>> > output of TCP_INFO in "some" tool below 250us.
>> >
>> > iperf has come a long way, but getting to where I could trust the
>> > largely "academic" codebase it was would take a ton of benchmarking.
>> >
>> > On Sat, Mar 27, 2021 at 2:09 PM Klatsky, Carl via Bloat
>> > <bloat@lists.bufferbloat.net> wrote:
>> > >
>> > > Toke,
>> > >
>> > > How you see this coming into Flent?  My understanding is that for latency & load test like RRUL:
>> > >
>> > > -iperf provides the bandwidth portion
>> > > -IRTT provides the latency portion, and if IRTT is not found, the test falls back to ICMP for the latency check
>> > >
>> > > Would netperf replace iperf for the bandwidth portion?  Do you see it being used for the latency portion?
>> > >
>> > > Thanks,
>> > > Carl
>> > >
>> > > -----Original Message-----
>> > > From: Bloat <bloat-bounces@lists.bufferbloat.net> On Behalf Of Toke Høiland-Jørgensen via Bloat
>> > > Sent: Friday, March 26, 2021 7:32 PM
>> > > To: bloat <bloat@lists.bufferbloat.net>
>> > > Subject: [Bloat] Netperf re-licensed as MIT
>> > >
>> > > Hopefully this means we can get it packaged for the distros that have thus far refused to because of the license - i.e., Debian and Fedora!
>> > >
>> > > -Toke
>> > > _______________________________________________
>> > > Bloat mailing list
>> > > Bloat@lists.bufferbloat.net
>> > > https://urldefense.com/v3/__https://lists.bufferbloat.net/listinfo/bloat__;!!CQl3mcHX2A!W_yWwnGSD6VuZvpn0BPv7Ta9GBP3f2dRWeJHPLS7RrrpBZ5gpyj5DZGnRYIfWVR5TXjR$
>> > > _______________________________________________
>> > > Bloat mailing list
>> > > Bloat@lists.bufferbloat.net
>> > > https://lists.bufferbloat.net/listinfo/bloat
>> >
>> >
>> >
>> > --
>> > "For a successful technology, reality must take precedence over public
>> > relations, for Mother Nature cannot be fooled" - Richard Feynman
>> >
>> > dave@taht.net <Dave Täht> CTO, TekLibre, LLC Tel: 1-831-435-0729
>>
>>
>>
>> --
>> "For a successful technology, reality must take precedence over public
>> relations, for Mother Nature cannot be fooled" - Richard Feynman
>>
>> dave@taht.net <Dave Täht> CTO, TekLibre, LLC Tel: 1-831-435-0729
>> _______________________________________________
>> Bloat mailing list
>> Bloat@lists.bufferbloat.net
>> https://lists.bufferbloat.net/listinfo/bloat
>
> --
> - Sent from my iPhone.



-- 
"For a successful technology, reality must take precedence over public
relations, for Mother Nature cannot be fooled" - Richard Feynman

dave@taht.net <Dave Täht> CTO, TekLibre, LLC Tel: 1-831-435-0729

Re: [Bloat] Netperf re-licensed as MIT

[Toke Høiland-Jørgensen]

Aaron Wood <woody77@gmail.com> writes:

> One of my long concerns with the RRUL test is that the ICMP ping test
> portion is not isochronous

That would be the UDP_RR test, you mean (ICMP is isochronous)? Yeah,
that is a bit annoying, but as Dave says if irtt is available, Flent
will use that, and that *is* isochronous :)

-Toke

Re: [Bloat] Netperf re-licensed as MIT

[Toke Høiland-Jørgensen]

Aaron Wood <woody77@gmail.com> writes:

> iperf3 isn’t “academic”, but is more focused on scientific computing (ESNet
> pushes a LOT of data CERN around, on 100Gbps backbones).
>
> But that also skews their usage/needs.  Very high throughput bulk transfers
> with long durations, over mixed systems. Not as many concerns about
> latency, except in that latency can cause messes with congestion
> control.

Yeah, I'm not too concerned about the code quality of iperf either - if
it ever reaches (rough) feature parity with netperf (list I posted
up-thread) I'm quite happy to turn it into an automatic fallback for
netperf in Flent, the same way we do with some of the other tools...

-Toke