서브메뉴
검색
Language Expressiveness Under Extreme Scarcity in Programmable Data Planes
Language Expressiveness Under Extreme Scarcity in Programmable Data Planes
Detailed Information
- 자료유형
- 학위논문 서양
- 최종처리일시
- 20250211151433
- ISBN
- 9798382807362
- DDC
- 004
- 저자명
- Hogan, Mary.
- 서명/저자
- Language Expressiveness Under Extreme Scarcity in Programmable Data Planes
- 발행사항
- [Sl] : Princeton University, 2024
- 발행사항
- Ann Arbor : ProQuest Dissertations & Theses, 2024
- 형태사항
- 127 p
- 주기사항
- Source: Dissertations Abstracts International, Volume: 85-12, Section: B.
- 주기사항
- Advisor: Rexford, Jennifer.
- 학위논문주기
- Thesis (Ph.D.)--Princeton University, 2024.
- 초록/해제
- 요약Today's networks must run a vast array of sophisticated applications that support services we rely on. These applications provide increased security (e.g., firewalls), better performance (e.g., congestion detection), and the ability to scale services (e.g., caching). The network devices that implement these applications need to operate at high speeds (100+Gbps) and be flexible enough to adapt applications to changing requirements. Programmable network devices have emerged as a way to customize network functionality, while guaranteeing high-speed processing. Writing applications for programmable devices, however, is notoriously difficult. Switches have a restrictive architecture to ensure line-rate processing, and their programming languages are very low-level. Programmers must understand how many resources (e.g., memory) each piece of their program requires. Deploying these applications often requires tedious optimization of their layout, with programmers manually writing, compiling, and testing an implementation, adjusting the design, and repeating.To better manage resource allocation, we present P4All, an extension of an existing programming language that allows programmers to define elastic data structures that stretch automatically to optimally use available resources. These structures are defined using symbolic values (that parameterize the size of the structure) and objective functions (that quantify the affect of size on performance). An optimization function specifies how to share resources amongst structures. We also create an optimizer that automatically finds the best resource allocation.There are many other choices programmers make beyond resource allocation, some of which likely depend on the expected workload. To automate those decisions, we present Parasol, a framework that allows programmers to define general, parameterized applications and automatically optimize their parameters. The parameters can represent a variety of implementation decisions, and may be optimized for high-level objectives defined by the programmer. Optimization is tailored to particular environments using a representative traffic trace.We implement a diverse set of applications in P4All and Parasol to evaluate the optimizers, and we compile the resulting optimized programs to an actual programmable switch. P4All and Parasol decouple programming languages from data-plane hardware to lift the burden of reasoning about low-level details from programmers and make it easier to develop applications.
- 일반주제명
- Computer science
- 일반주제명
- Computer engineering
- 일반주제명
- Systems science
- 키워드
- Programmers
- 기타저자
- Princeton University Computer Science
- 기본자료저록
- Dissertations Abstracts International. 85-12B.
- 전자적 위치 및 접속
- 로그인 후 원문을 볼 수 있습니다.
MARC
008250123s2024 us c eng d■001000017161702
■00520250211151433
■006m o d
■007cr#unu||||||||
■020 ▼a9798382807362
■035 ▼a(MiAaPQ)AAI31295432
■040 ▼aMiAaPQ▼cMiAaPQ
■0820 ▼a004
■1001 ▼aHogan, Mary.▼0(orcid)0000-0001-5915-5267
■24510▼aLanguage Expressiveness Under Extreme Scarcity in Programmable Data Planes
■260 ▼a[Sl]▼bPrinceton University▼c2024
■260 1▼aAnn Arbor▼bProQuest Dissertations & Theses▼c2024
■300 ▼a127 p
■500 ▼aSource: Dissertations Abstracts International, Volume: 85-12, Section: B.
■500 ▼aAdvisor: Rexford, Jennifer.
■5021 ▼aThesis (Ph.D.)--Princeton University, 2024.
■520 ▼aToday's networks must run a vast array of sophisticated applications that support services we rely on. These applications provide increased security (e.g., firewalls), better performance (e.g., congestion detection), and the ability to scale services (e.g., caching). The network devices that implement these applications need to operate at high speeds (100+Gbps) and be flexible enough to adapt applications to changing requirements. Programmable network devices have emerged as a way to customize network functionality, while guaranteeing high-speed processing. Writing applications for programmable devices, however, is notoriously difficult. Switches have a restrictive architecture to ensure line-rate processing, and their programming languages are very low-level. Programmers must understand how many resources (e.g., memory) each piece of their program requires. Deploying these applications often requires tedious optimization of their layout, with programmers manually writing, compiling, and testing an implementation, adjusting the design, and repeating.To better manage resource allocation, we present P4All, an extension of an existing programming language that allows programmers to define elastic data structures that stretch automatically to optimally use available resources. These structures are defined using symbolic values (that parameterize the size of the structure) and objective functions (that quantify the affect of size on performance). An optimization function specifies how to share resources amongst structures. We also create an optimizer that automatically finds the best resource allocation.There are many other choices programmers make beyond resource allocation, some of which likely depend on the expected workload. To automate those decisions, we present Parasol, a framework that allows programmers to define general, parameterized applications and automatically optimize their parameters. The parameters can represent a variety of implementation decisions, and may be optimized for high-level objectives defined by the programmer. Optimization is tailored to particular environments using a representative traffic trace.We implement a diverse set of applications in P4All and Parasol to evaluate the optimizers, and we compile the resulting optimized programs to an actual programmable switch. P4All and Parasol decouple programming languages from data-plane hardware to lift the burden of reasoning about low-level details from programmers and make it easier to develop applications.
■590 ▼aSchool code: 0181.
■650 4▼aComputer science
■650 4▼aComputer engineering
■650 4▼aSystems science
■653 ▼aComputer networks
■653 ▼aNetwork applications
■653 ▼aTraditional networks
■653 ▼aResource allocation
■653 ▼aProgrammers
■690 ▼a0984
■690 ▼a0464
■690 ▼a0790
■71020▼aPrinceton University▼bComputer Science.
■7730 ▼tDissertations Abstracts International▼g85-12B.
■790 ▼a0181
■791 ▼aPh.D.
■792 ▼a2024
■793 ▼aEnglish
■85640▼uhttp://www.riss.kr/pdu/ddodLink.do?id=T17161702▼nKERIS▼z이 자료의 원문은 한국교육학술정보원에서 제공합니다.
Preview
Export
ChatGPT Discussion
AI Recommended Related Books
Подробнее информация.
- Бронирование
- не существует
- моя папка
- Первый запрос зрения
- Non-Book Loan Application
- Nighttime Book Loan Application
Available after logging in.


