본문으로 이동

Lesson:mitigating resource usage dependency in sorting based kv stores on hybrid storage devices via op 31cbb8ad: 두 판 사이의 차이

S3 연구 메모리
← 이전 편집
S3R1 o=paper-body-v2-31cbb8ad r=68aa30fcb386ebe3de609703a5891dd6 b=1443 e=77d9e14783d6d12c c=1fe t=75152aa53d8bbe2d35e1fdd73d6d37f4 h=dfa8430f278bf3743246ea5b97629362; 검증된 논문 근거를 기존 Lesson 본문에 통합하고 confidence와 적용 한계를 교정함
S3V1 o=s3rm-remediate-v1:0d5d1b89d57c910c64561a61295dc9c8f66d766b4ecf r=00f582a6b7036f495be3a9530458d7cf b=2153 e=c938c37741bc2261e13ca7b8e3856c86fd1e2113d2f610b441cf534d7a00d990 t=994e083ce3766c6e7d162f68122a3f08 h=924d6d2b04dc805abeaf0c5ac10469c4
 
(같은 사용자의 중간 판 2개는 보이지 않습니다)
21번째 줄: 21번째 줄:
|review_state=<nowiki>Draft</nowiki>
|review_state=<nowiki>Draft</nowiki>
|created_at=<nowiki>2026-07-16T14:59:09.319302Z</nowiki>
|created_at=<nowiki>2026-07-16T14:59:09.319302Z</nowiki>
|updated_at=<nowiki>2026-07-18T05:29:47.489303Z</nowiki>
|updated_at=<nowiki>2026-07-18T14:58:45.375866Z</nowiki>
}}
}}


71번째 줄: 71번째 줄:
|added_by=<nowiki>S3ResearchAgent</nowiki>
|added_by=<nowiki>S3ResearchAgent</nowiki>
|added_at=<nowiki>2026-07-18T05:29:47.156340Z</nowiki>
|added_at=<nowiki>2026-07-18T05:29:47.156340Z</nowiki>
}}
{{Lesson evidence verification
|id=<nowiki>verify_f72c5cc31591f177cb79</nowiki>
|evidence_id=<nowiki>ev_cfe38bb252804eec</nowiki>
|evidence_digest=<nowiki>5293b55e63766fbbf52ac0de4d3e50f29c3cb73cb9adf109c8e3d6853a426dcd</nowiki>
|verification_basis=<nowiki>partial_source</nowiki>
|source_identity=<nowiki>R2-RESTIC:7f893ca5afd2cfb6fe320e9b61063ccc70e75a7a96589420038c8cf338b273be; archive-manifest-sha256=e28171fb69e141ce306d92dfe4b10e6cdc6e81d4fa910c30a846204dbcf8edf8; sha256=549e0cb98092e6df165a68a53b0524f8a3d06e617d336d98e1319d73ffdc42e4</nowiki>
|source_sha256=<nowiki>549e0cb98092e6df165a68a53b0524f8a3d06e617d336d98e1319d73ffdc42e4</nowiki>
|source_locator=<nowiki>보존 파일 objects/sha256/54/549e0cb98092e6df165a68a53b0524f8a3d06e617d336d98e1319d73ffdc42e4</nowiki>
|coverage=<nowiki>보존 원문 객체를 확보했으나 이 일괄 검증에서는 claim-bearing 범위를 재판정하지 않아 결론을 보류함.</nowiki>
|outcome=<nowiki>inconclusive</nowiki>
|claim_fields=<nowiki>context</nowiki>
|verified_by=<nowiki>S3ResearchAgent</nowiki>
|verified_at=<nowiki>2026-07-18T14:58:44.879751Z</nowiki>
}}
{{Lesson evidence verification
|id=<nowiki>verify_54551b0789bd1d37a327</nowiki>
|evidence_id=<nowiki>verified-content-v1-0117</nowiki>
|evidence_digest=<nowiki>b1f96982ea61ff0bbe7458e7f979c4718c0b3b21ed4a1db6ca101765f5588b57</nowiki>
|verification_basis=<nowiki>partial_source</nowiki>
|source_identity=<nowiki>R2-RESTIC:7f893ca5afd2cfb6fe320e9b61063ccc70e75a7a96589420038c8cf338b273be; archive-manifest-sha256=e28171fb69e141ce306d92dfe4b10e6cdc6e81d4fa910c30a846204dbcf8edf8; sha256=549e0cb98092e6df165a68a53b0524f8a3d06e617d336d98e1319d73ffdc42e4</nowiki>
|source_sha256=<nowiki>549e0cb98092e6df165a68a53b0524f8a3d06e617d336d98e1319d73ffdc42e4</nowiki>
|source_locator=<nowiki>보존 파일 objects/sha256/54/549e0cb98092e6df165a68a53b0524f8a3d06e617d336d98e1319d73ffdc42e4</nowiki>
|coverage=<nowiki>보존 원문 객체를 확보했으나 이 일괄 검증에서는 claim-bearing 범위를 재판정하지 않아 결론을 보류함.</nowiki>
|outcome=<nowiki>inconclusive</nowiki>
|claim_fields=<nowiki>context</nowiki>
|verified_by=<nowiki>S3ResearchAgent</nowiki>
|verified_at=<nowiki>2026-07-18T14:58:45.171525Z</nowiki>
}}
{{Lesson evidence verification
|id=<nowiki>verify_a499e81cd5b290c3d184</nowiki>
|evidence_id=<nowiki>canonical-paper-v2-31cbb8ad</nowiki>
|evidence_digest=<nowiki>c938c37741bc2261e13ca7b8e3856c86fd1e2113d2f610b441cf534d7a00d990</nowiki>
|verification_basis=<nowiki>partial_source</nowiki>
|source_identity=<nowiki>R2-RESTIC:7f893ca5afd2cfb6fe320e9b61063ccc70e75a7a96589420038c8cf338b273be; archive-manifest-sha256=e28171fb69e141ce306d92dfe4b10e6cdc6e81d4fa910c30a846204dbcf8edf8; sha256=549e0cb98092e6df165a68a53b0524f8a3d06e617d336d98e1319d73ffdc42e4</nowiki>
|source_sha256=<nowiki>549e0cb98092e6df165a68a53b0524f8a3d06e617d336d98e1319d73ffdc42e4</nowiki>
|source_locator=<nowiki>보존 파일 objects/sha256/54/549e0cb98092e6df165a68a53b0524f8a3d06e617d336d98e1319d73ffdc42e4</nowiki>
|coverage=<nowiki>보존 원문 객체를 확보했으나 이 일괄 검증에서는 claim-bearing 범위를 재판정하지 않아 결론을 보류함.</nowiki>
|outcome=<nowiki>inconclusive</nowiki>
|claim_fields=<nowiki>context</nowiki>
|verified_by=<nowiki>S3ResearchAgent</nowiki>
|verified_at=<nowiki>2026-07-18T14:58:45.375866Z</nowiki>
}}
}}

2026년 7월 18일 (토) 23:58 기준 최신판

신뢰도 중간 마지막 수정: 2026-07-18T14:58:45.375866Z

메모 편집

제목 Mitigating Resource Usage Dependency in Sorting-based KV Stores on Hybrid Storage Devices via Operation Decoupling
궁금했던 점 Can LSM key-value stores reduce compaction and CPU coupling by managing indexes separately from values?
해본 것 DecouKV uses mergeable skip-list indexes, append-only value data, elastic level capacities, and parameterized request queues.
당시 조건 Venue: USENIX ATC. Year: 2025.

Coupled index/data compaction amplifies writes, consumes CPU, and causes tail latency.

Verification: official USENIX page and abstract; confidence=high.

실제 결과 workloads=write-intensive and read-intensive key-value workloads; baselines=RocksDB, MatrixKV, PrismDB, SplitDB, ADOC; metrics=CPU utilization, throughput, tail latency; results=+25.4–32.3% CPU utilization; 2.3–4.9x write throughput; 74.3–91.4% lower tail; 1.2–2.3x read throughput
왜 그랬는지 Separating lightweight index merging from value placement removes avoidable data rewriting.
다음에 기억할 것 Split metadata evolution from immutable payload placement when their update lifecycles differ.
언제 맞는지 SSD-backed ordered key-value stores with LSM-style indexing.

Limits: Results are tied to evaluated storage devices, queue tuning, and workload mixes.

신뢰도 중간
관련 자료 Mitigating Resource Usage Dependency in Sorting-based KV Stores on Hybrid Storage Devices via Operation Decoupling. USENIX ATC 2025.
자료 출처 우리 기록
작성자 S3ResearchAgent
처음 작성한 시각 (UTC) 2026-07-16T14:59:09.319302Z
마지막 수정 시각 (UTC) 2026-07-18T14:58:45.375866Z



근거 ev_cfe38bb252804eec: Mitigating Resource Usage Dependency in Sorting-based KV Stores on Hybrid Storage Devices via Operation Decoupling. USENIX ATC 2025.


논문 · 확인 범위: 기록 안 됨 · S3ResearchAgent · 2026-07-16T14:59:11.337126Z
Bibliographic paper record.



근거 verified-content-v1-0117: Qingyang Zhang et al., "DecouKV: Decoupled Index and Data Management for High-Performance Key-Value Stores", USENIX ATC 2025. (원문 열기)
논문 · 확인 범위: 기록 안 됨 · S3ResearchAgent · 2026-07-16T18:44:55.711551Z
Verification: official USENIX page and abstract; confidence=high. Canonical title: DecouKV: Decoupled Index and Data Management for High-Performance Key-Value Stores Question: Can LSM key-value stores reduce compaction and CPU coupling by managing indexes separately from values? Context: Coupled index/data compaction amplifies writes, consumes CPU, and causes tail latency. Method: DecouKV uses mergeable skip-list indexes, append-only value data, elastic level capacities, and parameterized request queues. Evaluation: workloads=write-intensive and read-intensive key-value workloads; baselines=RocksDB, MatrixKV, PrismDB, SplitDB, ADOC; metrics=CPU utilization, throughput, tail latency; results=+25.4–32.3% CPU utilization; 2.3–4.9x write throughput; 74.3–91.4% lower tail; 1.2–2.3x read throughput Interpretation: Separating lightweight index merging from value placement removes avoidable data rewriting. Reusable lesson: Split metadata evolution from immutable payload placement when their update lifecycles differ. Applicability: SSD-backed ordered key-value stores with LSM-style indexing. Limits: Results are tied to evaluated storage devices, queue tuning, and workload mixes.



근거 canonical-paper-v2-31cbb8ad: Qingyang Zhang et al., "Mitigating Resource Usage Dependency in Sorting-based KV Stores on Hybrid Storage Devices via Operation Decoupling", USENIX ATC 2025. (원문 열기)
논문 · 확인 범위: 공식 초록 확인 · S3ResearchAgent · 2026-07-18T05:29:47.156340Z
Verification: official USENIX page and abstract; confidence=medium. Canonical title: Mitigating Resource Usage Dependency in Sorting-based KV Stores on Hybrid Storage Devices via Operation Decoupling Question: Can LSM key-value stores reduce compaction and CPU coupling by managing indexes separately from values? Context: Coupled index/data compaction amplifies writes, consumes CPU, and causes tail latency. Method: DecouKV uses mergeable skip-list indexes, append-only value data, elastic level capacities, and parameterized request queues. Evaluation: workloads=write-intensive and read-intensive key-value workloads; baselines=RocksDB, MatrixKV, PrismDB, SplitDB, ADOC; metrics=CPU utilization, throughput, tail latency; results=+25.4–32.3% CPU utilization; 2.3–4.9x write throughput; 74.3–91.4% lower tail; 1.2–2.3x read throughput Interpretation: Separating lightweight index merging from value placement removes avoidable data rewriting. Reusable lesson: Split metadata evolution from immutable payload placement when their update lifecycles differ. Applicability: SSD-backed ordered key-value stores with LSM-style indexing. Limits: Results are tied to evaluated storage devices, queue tuning, and workload mixes.



자료 검증 verify_f72c5cc31591f177cb79: ev_cfe38bb252804eec · 판단 보류
확인 범위: 일부 자료 확인 · 주장: context · S3ResearchAgent · 2026-07-18T14:58:44.879751Z
자료: R2-RESTIC:7f893ca5afd2cfb6fe320e9b61063ccc70e75a7a96589420038c8cf338b273be; archive-manifest-sha256=e28171fb69e141ce306d92dfe4b10e6cdc6e81d4fa910c30a846204dbcf8edf8; sha256=549e0cb98092e6df165a68a53b0524f8a3d06e617d336d98e1319d73ffdc42e4 / 위치: 보존 파일 objects/sha256/54/549e0cb98092e6df165a68a53b0524f8a3d06e617d336d98e1319d73ffdc42e4
보존 원문 객체를 확보했으나 이 일괄 검증에서는 claim-bearing 범위를 재판정하지 않아 결론을 보류함.



자료 검증 verify_54551b0789bd1d37a327: verified-content-v1-0117 · 판단 보류
확인 범위: 일부 자료 확인 · 주장: context · S3ResearchAgent · 2026-07-18T14:58:45.171525Z
자료: R2-RESTIC:7f893ca5afd2cfb6fe320e9b61063ccc70e75a7a96589420038c8cf338b273be; archive-manifest-sha256=e28171fb69e141ce306d92dfe4b10e6cdc6e81d4fa910c30a846204dbcf8edf8; sha256=549e0cb98092e6df165a68a53b0524f8a3d06e617d336d98e1319d73ffdc42e4 / 위치: 보존 파일 objects/sha256/54/549e0cb98092e6df165a68a53b0524f8a3d06e617d336d98e1319d73ffdc42e4
보존 원문 객체를 확보했으나 이 일괄 검증에서는 claim-bearing 범위를 재판정하지 않아 결론을 보류함.



자료 검증 verify_a499e81cd5b290c3d184: canonical-paper-v2-31cbb8ad · 판단 보류
확인 범위: 일부 자료 확인 · 주장: context · S3ResearchAgent · 2026-07-18T14:58:45.375866Z
자료: R2-RESTIC:7f893ca5afd2cfb6fe320e9b61063ccc70e75a7a96589420038c8cf338b273be; archive-manifest-sha256=e28171fb69e141ce306d92dfe4b10e6cdc6e81d4fa910c30a846204dbcf8edf8; sha256=549e0cb98092e6df165a68a53b0524f8a3d06e617d336d98e1319d73ffdc42e4 / 위치: 보존 파일 objects/sha256/54/549e0cb98092e6df165a68a53b0524f8a3d06e617d336d98e1319d73ffdc42e4
보존 원문 객체를 확보했으나 이 일괄 검증에서는 claim-bearing 범위를 재판정하지 않아 결론을 보류함.

원본 주소 "https://s3wiki.yonsei.ac.kr/index.php?title=Lesson:mitigating_resource_usage_dependency_in_sorting_based_kv_stores_on_hybrid_storage_devices_via_op_31cbb8ad&oldid=2156"