[P4] cikk váz

[Sándor Laki]

Hi,

Átírtam a szerkezetet. Amúgy ha mindenhova tudunk értelmes dolgokat 
írni, akkor szerintem nem is lesz rossz. Az idő a kérdés, továbbá, hogy 
nincs másik hardware-re HAL. Akinek van kis ideje, az írjon részeket, 
fejlődjön!!! A címmel és absztrakttal beregisztrálom a cikket, ezt 
nézzétek meg:

\title{OpenP4C: A Multi-Target Compiler for High Performance Protocol 
Independent Packet Processors}

\begin{abstract}
Though the programmability of the control plane has been thoroughly 
examined and developed in the past years, only a limited number of 
studies go beyond the consideration that the data plane is only a 
collection of simple packet forwarding devices with the support of 
existing protocols. OpenFlow, a popular example, is a very expressive 
data plane programming language, but it is still restricted to the 
support of a subset of existing protocol headers, inhibiting the 
introduction of new data plane protocols. Furthermore, its 
expressiveness comes with a severe performance price caused by the 
excessive packet classification the switches must do in the fast path. 
To overcome these limitations, new data plane programming models like P4 
have emerged recently. P4 is a high level language for programming 
network switches that allows for great flexibility in the description of 
packet structure and processing, independent of the specifics of the 
underlying hardware. In this paper, we introduce \textsc{OpenP4C} a 
multi-target compiler that generates high performance and target 
independent switch program from a P4 description. To support multiple 
targets, the compiler generates a core switch code relying on an 
abstraction layer of the underlying hardware that should be implemented 
for the given target architecture. To avoid performance degradation, the 
boundaries of this separation should carefully  be chosen, since the 
core program only responsible for target-independent optimization, while 
the implementation of the abstraction layer should cover 
target-dependent performance enhancement. To analyze the performance of 
the proposed compiler, thorough measurements have been carried out in 
Intel platform, considering different use case scenarios. Finally, we 
also show that the switch generated by \textsc{OpenP4C} can easily scale 
beyond 100 Gbps even in a singe Intel blade platform, outperforming 
\textsc{OpenVSwitch} in similar scenarios.
\end{abstract}

Üdv.
Sanyi


On 2016.07.18. 21:45, Tejfel Máté wrote:
> Sziasztok!
>
>   Felraktam egy nagyon, nagyon kezdetleges vázat, hogy miről lehetne 
> írni a cikkben a délelőtti beszélgetés alapján... Nyugodtan lehet 
> javítani + feltölteni tartalommal...
>
>               M.
> _______________________________________________
> P4 mailing list
> P4 at plc.inf.elte.hu
> https://plc.inf.elte.hu/mailman/listinfo/p4

-- 
Sándor Laki, PhD
Assistant professor
Department of Information Systems
Eötvös Loránd University
Pázmány Péter stny. 1/C
H-1117, Budapest, Hungary
Room 2.506
Web: http://lakis.web.elte.hu
Phone: +36 1 372 2869 / 8477
Cell: +36 70 374 2646


---
Ezt az e-mailt az Avast víruskereső szoftver átvizsgálta.
https://www.avast.com/antivirus