[v6ops] 464XLAT tester -- Re: Looking for free software MAP-T (BR) implementations

Dear Henri,

9/3/2023 3:54 PM keltezéssel, Henri Alves de Godoy írta:
> Congrats on your student's work.

Thank you very much for that. :-)

Maptperf is a free software and we would be happy if you could use it. 
It is documented in Ahmed's paper which is currently under review. For 
this reason, I cannot share his paper publicly yet, but we can send its 
review version to you in a private e-mail, if you are interested in it.

> Do you have any plans to do the same tester in 464xlat?

No, I don't, because I have already done it. ;-)

As CLAT implements stateless NAT46, also called SIIT, it was the 
original target of my siitperf: https://github.com/lencsegabor/siitperf
Its original version, which literally followed the fixed test frame 
format of RFC 2544, is documented in this paper: 
https://doi.org/10.1587/transcom.2019EBN0010
Its extension to use RFC 4814 pseudorandom port numbers is documented 
here: https://doi.org/10.11601/ijates.v9i3.291

As PLAT is a stateful NAT64 gateway, it can be benchmarked according to 
our methodology for stateful NATxy gateways: 
https://datatracker.ietf.org/doc/html/draft-ietf-bmwg-benchmarking-stateful
I have extended siitperf for stateful NAT44 / NAT64 tests. It is 
documented in: https://doi.org/10.1016/j.comcom.2022.05.028
And our methodology has been validated by benchmarking three radically 
different stateful NAT64 implementations (Jool, tayga+iptables, OpenBSD 
PF) in: https://doi.org/10.1016/j.comcom.2023.08.009

I hope that the above documentation can help you if you would like to 
use siitperf.

However, there is one limitation: currently it runs only under Debian 9. 
(It can be compiled with its DPDK version. I plan to upgrade it to the 
current DPDK version in a few months.)

>
> The draft that you are building is important and great for having 
> parameters for using existing translation tools.
Thank you!
> If I can contribute something, let me know.
>
It would be a great help for me if you could review the scalability 
draft ( 
https://datatracker.ietf.org/doc/html/draft-lencse-v6ops-transition-scalability 
) and send the results to the v6ops mailing list. Your comments can help 
me to make the document more useful. E.g., What is missing? What is 
difficult to understand?

(The draft has been expired, but I want to continue it with further 
results. As I am an academic researcher, I first need to publish our 
results in peer reviewed journals, and then I can add them to the draft 
after that. Of course, I have ideas, what and how to measure, and 
perhaps my ideas are good enough to publish the results in journal 
papers, but I need input from network operators to be able to produce 
the kinds of results what they need.)

Best regards,

Gábor

> Att,
> Henri
>
> Em sex., 1 de set. de 2023 às 14:28, Gábor LENCSE <lencse@hit.bme.hu> 
> escreveu:
>
>     Dear All,
>
>     One of my PhD students has implemented and RFC 8219 compliant
>     MAP-T BR
>     tester called Maptperf:
>     https://github.com/alhamadani-ahmed/Maptperf_v1.0
>
>     He plans to benchmark various free software MAP-T BR implementations.
>     However, currently we know only one such software, Jool:
>     https://www.jool.mx/en/run-mapt.html
>
>     We are looking for contemporary software, which can be compiled on
>     modern Linux OS-es (e.g. Debian 11 or 12, Ubuntu 22.x).
>
>     Could you please suggest a few?
>
>     We plan to add the summary of the results (after publication of
>     all the
>     details in a good reputation journal as an open access paper) to
>     my I-D
>     "Scalability of IPv6 Transition Technologies for IPv4aaS":
>     https://datatracker.ietf.org/doc/html/draft-lencse-v6ops-transition-scalability
>
>     Thank you very much for your help in advance!
>
>     Best regards,
>
>     Gábor
>
>     _______________________________________________
>     v6ops mailing list
>     v6ops@ietf.org
>     https://www.ietf.org/mailman/listinfo/v6ops
>
>
>
> -- 
>