]> 2026-09-08T15:37:43+09:00 Lesson:phasedrr read reclaim scheduling without page level access counting 69578f95 3000 ko TLC/QLC SSD FTL read-reclaim. Limits: SSDsim 기반이며 세 window·threshold와 trace hotness에 민감하다. 세 시간창의 working set 교집합으로 hottest/hotter/tepid를 구분해 hard threshold 전에 단계적으로 이동한다. S3ResearchAgent high Venue: MSST. Year: 2024. 기존 hot-page 선이동은 DRAM의 page counter가 캐시 공간을 잠식한다. Verification: full_text; confidence=high. 2026-07-16T14:56:05Z 2461238.1222801 PhasedRR: Read Reclaim Scheduling without Page-level Access Counting. MSST 2024. 정확한 빈도 카운트 대신 최근성 집합의 반복 출현만으로 충분한 hotness 계층을 만들 수 있다. PhasedRR: Read Reclaim Scheduling without Page-level Access Counting workloads=SSDsim; realistic disk traces; ARM Cortex-A7 controller model; baselines=RL-RR; Reallocation; existing RR scheduling; metrics=overall I/O latency; read latency; erase operations; DRAM overhead; results=Overall I/O latency -31.3%; read latency -33.6%; erases -6.1% on average. 페이지별 read counter 없이 SSD read-reclaim 지연을 줄일 수 있는가? lab 메타데이터 비용이 크면 exact counter를 windowed set membership으로 대체하라. Draft 2026-07-18T14:58:52Z 2461240.124213 2026-07-18T14:58:52Z 2461240.124213 phasedrr read reclaim scheduling without page level access counting 69578f95 3000 ev_d70efcd9ba7a4f5d 2026-07-16T14:56:07Z 2461238.1223032 S3ResearchAgent PhasedRR: Read Reclaim Scheduling without Page-level Access Counting. MSST 2024. paper Bibliographic paper record. not_recorded phasedrr read reclaim scheduling without page level access counting 69578f95# 03104df6f98d1865ea8e64fb7a2e80b9 3000 verified-content-v1-0049 2026-07-16T18:53:47Z 2461238.2873495 S3ResearchAgent Li Cai et al., "PhasedRR: Read Reclaim Scheduling without Page-level Access Counting", MSST 2024. paper Verification: full_text; confidence=high. Canonical title: PhasedRR: Read Reclaim Scheduling without Page-level Access Counting Question: 페이지별 read counter 없이 SSD read-reclaim 지연을 줄일 수 있는가? Context: 기존 hot-page 선이동은 DRAM의 page counter가 캐시 공간을 잠식한다. Method: 세 시간창의 working set 교집합으로 hottest/hotter/tepid를 구분해 hard threshold 전에 단계적으로 이동한다. Evaluation: workloads=SSDsim; realistic disk traces; ARM Cortex-A7 controller model; baselines=RL-RR; Reallocation; existing RR scheduling; metrics=overall I/O latency; read latency; erase operations; DRAM overhead; results=Overall I/O latency -31.3%; read latency -33.6%; erases -6.1% on average. Interpretation: 정확한 빈도 카운트 대신 최근성 집합의 반복 출현만으로 충분한 hotness 계층을 만들 수 있다. Reusable lesson: 메타데이터 비용이 크면 exact counter를 windowed set membership으로 대체하라. Applicability: TLC/QLC SSD FTL read-reclaim. Limits: SSDsim 기반이며 세 window·threshold와 trace hotness에 민감하다. not_recorded phasedrr read reclaim scheduling without page level access counting 69578f95# 737b1c212b4990dd1b97365862d4d425 3000 canonical-paper-v2-69578f95 2026-07-18T05:33:54Z 2461239.731875 S3ResearchAgent Li Cai et al., "PhasedRR: Read Reclaim Scheduling without Page-level Access Counting", MSST 2024. paper Verification: full_text; confidence=high. Canonical title: PhasedRR: Read Reclaim Scheduling without Page-level Access Counting Question: 페이지별 read counter 없이 SSD read-reclaim 지연을 줄일 수 있는가? Context: 기존 hot-page 선이동은 DRAM의 page counter가 캐시 공간을 잠식한다. Method: 세 시간창의 working set 교집합으로 hottest/hotter/tepid를 구분해 hard threshold 전에 단계적으로 이동한다. Evaluation: workloads=SSDsim; realistic disk traces; ARM Cortex-A7 controller model; baselines=RL-RR; Reallocation; existing RR scheduling; metrics=overall I/O latency; read latency; erase operations; DRAM overhead; results=Overall I/O latency -31.3%; read latency -33.6%; erases -6.1% on average. Interpretation: 정확한 빈도 카운트 대신 최근성 집합의 반복 출현만으로 충분한 hotness 계층을 만들 수 있다. Reusable lesson: 메타데이터 비용이 크면 exact counter를 windowed set membership으로 대체하라. Applicability: TLC/QLC SSD FTL read-reclaim. Limits: SSDsim 기반이며 세 window·threshold와 trace hotness에 민감하다. full_text phasedrr read reclaim scheduling without page level access counting 69578f95# 2712015fd2bf894ccc23f80ec9f20840 3000 2026-07-18T14:58:52Z 2461240.124213 S3ResearchAgent verify_a662e63e00bdbd6caac4 partial_source context 보존 원문 객체를 확보했으나 이 일괄 검증에서는 claim-bearing 범위를 재판정하지 않아 결론을 보류함. inconclusive d7d1f39acf07dfe63576cc585e911488f880cd0b50e7441fdd3cda02220b2523 R2-RESTIC:7f893ca5afd2cfb6fe320e9b61063ccc70e75a7a96589420038c8cf338b273be; archive-manifest-sha256=e28171fb69e141ce306d92dfe4b10e6cdc6e81d4fa910c30a846204dbcf8edf8; sha256=d7d1f39acf07dfe63576cc585e911488f880cd0b50e7441fdd3cda02220b2523 보존 파일 objects/sha256/d7/d7d1f39acf07dfe63576cc585e911488f880cd0b50e7441fdd3cda02220b2523 ev_d70efcd9ba7a4f5d 05bb61e6db497b76c2a502e03ed964ed99113b56e64368dbd65ee87977885cee phasedrr read reclaim scheduling without page level access counting 69578f95# 8f70fb56b393ba2a9e85459128eb7e76 3000 2026-07-18T14:58:52Z 2461240.124213 S3ResearchAgent verify_4b998fc5c80f46447e51 partial_source context 보존 원문 객체를 확보했으나 이 일괄 검증에서는 claim-bearing 범위를 재판정하지 않아 결론을 보류함. inconclusive d7d1f39acf07dfe63576cc585e911488f880cd0b50e7441fdd3cda02220b2523 R2-RESTIC:7f893ca5afd2cfb6fe320e9b61063ccc70e75a7a96589420038c8cf338b273be; archive-manifest-sha256=e28171fb69e141ce306d92dfe4b10e6cdc6e81d4fa910c30a846204dbcf8edf8; sha256=d7d1f39acf07dfe63576cc585e911488f880cd0b50e7441fdd3cda02220b2523 보존 파일 objects/sha256/d7/d7d1f39acf07dfe63576cc585e911488f880cd0b50e7441fdd3cda02220b2523 verified-content-v1-0049 2b68dc3279073ec4fad4c3ad47273616061143f8090209bf917a36d519a41619 phasedrr read reclaim scheduling without page level access counting 69578f95# 69c9c77cb2a24ecadb2d1dd5a2954b0a 3000 2026-07-18T14:58:52Z 2461240.124213 S3ResearchAgent verify_c96733bdd3554060b7a3 full_text observation,interpretation,reusable_lesson observation=supported; interpretation=supported; reusable_lesson=supported supports d7d1f39acf07dfe63576cc585e911488f880cd0b50e7441fdd3cda02220b2523 R2-RESTIC:7f893ca5afd2cfb6fe320e9b61063ccc70e75a7a96589420038c8cf338b273be; archive-manifest-sha256=e28171fb69e141ce306d92dfe4b10e6cdc6e81d4fa910c30a846204dbcf8edf8; sha256=d7d1f39acf07dfe63576cc585e911488f880cd0b50e7441fdd3cda02220b2523; independently adjudicated claim-bearing primary source PDF pp. 5–9, §IV Experimental Evaluation, Table I and Figures 6–11; design §III. canonical-paper-v2-69578f95 eb7982dd7255587e2a10cf1fb9e9b19d8a4223100b90e7313c01af30e2301be1 phasedrr read reclaim scheduling without page level access counting 69578f95# 8e978ac1e0227dd96cc884ca3fa7ba48