메뉴 건너뛰기
.. 내서재 .. 알림
소속 기관/학교 인증
인증하면 논문, 학술자료 등을  무료로 열람할 수 있어요.
한국대학교, 누리자동차, 시립도서관 등 나의 기관을 확인해보세요
(국내 대학 90% 이상 구독 중)
로그인 회원가입 고객센터 ENG
주제분류

추천
검색
질문

논문 기본 정보

자료유형
학술저널
저자정보
Eunji Lee (Soongsil University)
저널정보
Korean Institute of Information Scientists and Engineers Journal of Computing Science and Engineering Journal of Computing Science and Engineering Vol.13 No.4
발행연도
2019.12
수록면
141 - 150 (10page)
DOI
10.5626/JCSE.2019.13.4.141

이용수

표지
📌
연구주제
📖
연구배경
🔬
연구방법
🏆
연구결과
AI에게 요청하기
추천
검색
질문

초록· 키워드

오류제보하기
The underlying technology trend stresses the design of the software. With increasing use of many-core computers that equip a large number of independent processor units, enhancing scalability and concurrency of commercial software is of crucial importance. To fulfill this demand, non-blocking implementations of the popular data structures are extensively explored both in academia and industry to effectively harness the massive parallelism in a many-core system. This paper presents a new non-blocking skip list that is not only scalable but also provides low latency even under high-concurrency pressure. Existing techniques for parallelizing skip lists rely on retrying insertion operations which fail because one thread interferes with another. This approach can introduce long-tail latency when multiple threads compete for access to the same links. We address this issue by exploiting the probabilistic nature of the skip list by allowing insertion operations to terminate after a failure, even if all the links from a node have not been updated. The resulting reduction in the heights of many nodes changes the statistical properties of the links, on which the efficiency of the skip list depends. We compensate this side-effect by recording reductions in node height and recompense for them when new nodes are created. To demonstrate the effectiveness of our approach, we implement a prototype of our low-latency non-blocking skip list, and the measurement study with various workloads shows that our skip list provides more scalable performance and lower tail latency compared to existing skip lists.

목차

Abstract
I. INTRODUCTION
II. THE SKIP LIST
III. A LOW-LATENCY NON-BLOCKING SKIP LIST WITH RETRIAL-FREE SYNCHRONIZATION
IV. PERFORMANCE EVALUATION
V. RELATED WORK
VI. CONCLUSION
REFERENCES

참고문헌 (38)

참고문헌 신청

함께 읽어보면 좋을 논문

논문 유사도에 따라 DBpia 가 추천하는 논문입니다. 함께 보면 좋을 연관 논문을 확인해보세요!

이 논문의 저자 정보

최근 본 자료

전체보기

댓글(0)

0