Novel contiguity transformed writes and adaptive strip prefetched reads for the high performance of RAID = 레이드의 고성능을 위한 새로운 연속성 변형에 의한 쓰기와 적응적 조각 프리페치에 의한 읽기

Cited 0 time in webofscience Cited 0 time in scopus
  • Hit : 487
  • Download : 0
It is very important to improve the read and write performance of RAID. The write performance of RAID is related to disk scheduling algorithms and destage algorithms. The read performance of RAID is associated with prefetching and caching technologies. This thesis presents a destage algorithm for fragmented writes in RAID, prefetching algorithms that resolves the fundamental problems of sequential prefetching when it confronts with striped disk arrays, and a sophisticated prefetching and buffer management algorithm that resolves five problems that are newly addressed in this thesis. Contiguous reads and writes outperform fragmented reads and writes even though the starting positions and the end positions of the kinds of I/O are the same, fragmented reads and writes are forcefully transformed into contiguous reads and writes via a proposed matrix-stripe-cache-based contiguity transform (MSC-CT) method, which employs a rule of consistency for data integrity at the block level, and a rule of performance that ensures no performance degradation. MSC-CT performs at destage in a redundant rray of independent disks (RAID) array. Traditional studies on disk arrays have focused on parallelism or load balance of disks. However, this thesis reveals that the independency of disks is more important than parallelism for concurrent reads of larger numbers of processes in striped disk arrays, whereas parallelism is only significant for concurrent reads of small numbers of processes. Hence, this thesis propose two types of sequential prefetching: strip-aligned sequential prefetching (SASEQP) for independency and massive stripe prefetching (MSP) for parallelism. Conventional prefetching schemes regard prediction accuracy as important because useless data prefetched by a faulty prediction may pollute the cache. If prefetching requires considerably low read cost but the prediction is not accurate, it may or may not be beneficial depending on the situation. ...
Park, kyu-Horesearcher박규호researcher
한국과학기술원 : 전기및전자공학전공,
Issue Date
303631/325007  / 020055838

학위논문(박사) - 한국과학기술원 : 전기및전자공학전공, 2008. 8., [ xii,124 p. ]


storage; operating system; RAID; I/O; cache; 저장장치; 운영체제; 레이드; 입출력; 캐시; storage; operating system; RAID; I/O; cache; 저장장치; 운영체제; 레이드; 입출력; 캐시

Appears in Collection
Files in This Item
There are no files associated with this item.


  • mendeley


rss_1.0 rss_2.0 atom_1.0