속성:Evidence note
자료에서 확인할 수 있는 점이나 한계를 적습니다.
p
Verification: official_abstract; confidence=medium.
Canonical title: Polymorphic Error Correction
Question: Can one redundancy budget flexibly protect memory against different fault models while also providing strong integrity authentication?
Context: Fixed ECC allocates check bits to one correction profile; secure memory also needs a MAC within tight metadata budgets.
Method: Polymorphic Error Correction reinterprets the same redundancy across fault models, combining an inline MAC with iterative correction.
Evaluation: workloads=64-byte cache lines; 40-bit DDR5 channels; multiple memory fault models; baselines=fixed ECC plus MAC designs; metrics=fault correction; detection probability; MAC width; results=near-100% detection; up to 60-bit MAC; supports multiple correction modes
Interpretation: Redundancy can be encoded for multiple operating points instead of hard-wiring one reliability/security tradeoff.
Reusable lesson: Make protection metadata polymorphic so systems can adapt correction strength and authentication without changing storage overhead.
Applicability: Secure DDR5 memory systems combining ECC and integrity checks.
Limits: Assumes a MAC+ECC secure-memory setting; performance, iteration latency, and all fault assumptions were not verified from full text. +
Source: Yonsei University Computer Systems Laboratory publication list supplied by the user. Manifestation 1 of 1. +
정본 Lesson 보강 근거. 검토 원본: Lesson:technical_review_power_consumption_prediction_and_power_aware_packing_in_consolidated_environmen_58907b8a. 확인 범위: official_abstract. 확인한 자료: https://doi.org/10.1109/TC.2010.91 ; https://yonsei.elsevierpure.com/en/publications/power-consumption-prediction-and-power-aware-packing-in-consolida/ ; https://pure.psu.edu/en/publications/power-consumption-prediction-and-power-aware-packing-in-consolida. 질문, 방법, 평가, 해석, 재사용 교훈, 적용 범위와 한계를 같은 Lesson 본문에 통합했습니다. +
Bibliographic paper record. +
Verification: official_abstract; confidence=high.
Question: How can future DRAM-system reliability be predicted when technology and protection mechanisms change?
Context: Device-level aggregate rates do not reveal which internal DRAM component failed or how a new organization/protection scheme changes outcomes.
Method: The work derives an empirical component-level fault model from public field-error logs and projects it to DDR5, HBM3, and LPDDR5 with on-die ECC and repair options.
Evaluation: workloads=public memory-error logs; projected DDR5, HBM3, and LPDDR5 systems; baselines=alternative on-die ECC and repair configurations; metrics=predicted system reliability; fault coverage; results=no quantitative headline stated in accessible primary abstract
Interpretation: Mapping observed errors to physical components enables architecture-specific projections instead of reusing aggregate failure rates.
Reusable lesson: Model failures at the component granularity affected by the design decision being evaluated.
Applicability: DRAM architecture, ECC, repair, and future-system reliability planning.
Limits: Predictions extrapolate public production logs through assumed component mappings and future-device parameters. +
Verification: official_abstract; confidence=medium.
Question: How can future DRAM-system reliability be predicted when technology and protection mechanisms change?
Context: Device-level aggregate rates do not reveal which internal DRAM component failed or how a new organization/protection scheme changes outcomes.
Method: The work derives an empirical component-level fault model from public field-error logs and projects it to DDR5, HBM3, and LPDDR5 with on-die ECC and repair options.
Evaluation: workloads=public memory-error logs; projected DDR5, HBM3, and LPDDR5 systems; baselines=alternative on-die ECC and repair configurations; metrics=predicted system reliability; fault coverage; results=no quantitative headline stated in accessible primary abstract
Interpretation: Mapping observed errors to physical components enables architecture-specific projections instead of reusing aggregate failure rates.
Reusable lesson: Model failures at the component granularity affected by the design decision being evaluated.
Applicability: DRAM architecture, ECC, repair, and future-system reliability planning.
Limits: Predictions extrapolate public production logs through assumed component mappings and future-device parameters. +
Verification: abstract_only; confidence=medium.
Canonical title: Prism: Optimizing Key-Value Store for Modern Heterogeneous Storage Devices
Question: 지연·대역폭 특성이 다른 영속 메모리와 NVMe를 KV 저장소에서 함께 활용할 수 있는가?
Context: 단일 계층 정책은 Optane DCPMM과 NVMe의 상이한 장점을 놓친다.
Method: 읽기 지연·대역폭을 균형 배치하고 계층 간 병렬성과 crash consistency를 제공한다.
Evaluation: workloads=KV workloads on Optane DCPMM and NVMe; baselines=state-of-the-art KV stores; metrics=throughput; tail latency; results=Up to 13.1× throughput.
Interpretation: 이질 저장장치에서는 데이터 배치와 동시성·복구를 공동 설계해야 한다.
Reusable lesson: 매체별 역할을 고정하지 말고 병목 지표에 따라 균형화하라.
Applicability: PM+SSD 계층형 KV 저장소.
Limits: 현재 단종된 Optane 계열 하드웨어 의존성이 크고 초록에 세부 baseline이 없다. +
Verification: abstract_only; confidence=high.
Canonical title: Prism: Optimizing Key-Value Store for Modern Heterogeneous Storage Devices
Question: 지연·대역폭 특성이 다른 영속 메모리와 NVMe를 KV 저장소에서 함께 활용할 수 있는가?
Context: 단일 계층 정책은 Optane DCPMM과 NVMe의 상이한 장점을 놓친다.
Method: 읽기 지연·대역폭을 균형 배치하고 계층 간 병렬성과 crash consistency를 제공한다.
Evaluation: workloads=KV workloads on Optane DCPMM and NVMe; baselines=state-of-the-art KV stores; metrics=throughput; tail latency; results=Up to 13.1× throughput.
Interpretation: 이질 저장장치에서는 데이터 배치와 동시성·복구를 공동 설계해야 한다.
Reusable lesson: 매체별 역할을 고정하지 말고 병목 지표에 따라 균형화하라.
Applicability: PM+SSD 계층형 KV 저장소.
Limits: 현재 단종된 Optane 계열 하드웨어 의존성이 크고 초록에 세부 baseline이 없다. +
Bibliographic paper record. +
Verification status (2026-07-17): no public primary paper, preprint, thesis, DOI, proceedings entry, or project repository was located after exact-title/distinctive-phrase searches across ACM DL, IEEE Xplore, USENIX, VLDB/DBLP, arXiv, and the relevant official lab publication pages.
Canonical title: PruneOff: Offloading LSM Read-Path Pruning in Disaggregated Key-Value Stores.
Publication metadata: authors, venue, year, DOI, and canonical paper URL remain unverified.
Research content: question beyond the title, method, implementation, workloads, baselines, metrics, quantitative results, interpretation, applicability, and limitations are unavailable from a public primary source and are intentionally not inferred.
Status: treat as an unpublished/private work-in-progress candidate, not as a content-verified published paper. Absence from public indexes does not prove that the work does not exist.
Follow-up: replace this status note with a paper/thesis/preprint evidence record if a public primary source becomes available. +
Verification: unresolved; title-only search with no public primary source found; confidence=low.
Canonical title: PruneOff: Offloading LSM Read-Path Pruning in Disaggregated Key-Value Stores
Question:
Context:
Method:
Evaluation: workloads=; baselines=; metrics=; results=
Interpretation:
Reusable lesson:
Applicability:
Limits: No publicly accessible primary-source paper, publisher record, DOI, author manuscript, or official program entry was located; content and publication status cannot be verified. +
Verification: unresolved; title-only search with no public primary source found; confidence=low.
Canonical title: PruneOff: Offloading LSM Read-Path Pruning in Disaggregated Key-Value Stores
Question:
Context:
Method:
Evaluation: workloads=; baselines=; metrics=; results=
Interpretation:
Reusable lesson:
Applicability:
Limits: No publicly accessible primary-source paper, publisher record, DOI, author manuscript, or official program entry was located; content and publication status cannot be verified. +
Bibliographic paper record. +
Verification: abstract_only; confidence=medium.
Canonical title: PVM: Efficient Shadow Paging for Deploying Secure Containers in Cloud-native Environment
Question: 하드웨어 가상화 기능을 게스트에 노출하지 않고 안전 컨테이너의 메모리 가상화를 빠르게 할 수 있는가?
Context: nested virtualization은 복잡·비싸고 전통 shadow paging은 동시성·일관성 병목이 있다.
Method: PVM은 게스트 하이퍼바이저와 최소 공유 영역, 효율적 shadow paging을 사용한다.
Evaluation: workloads=Alibaba/Ant production, tens of thousands containers/day; CPU, memory, I/O benchmarks; baselines=nested KVM; metrics=memory virtualization performance; CPU/I/O performance; results=Outperforms nested KVM for concurrent memory virtualization; exact aggregate not abstract-verified.
Interpretation: 보안 경계를 유지하면서 게스트에 필요한 최소 협업 인터페이스만 노출할 수 있다.
Reusable lesson: nested 계층은 최소 공유 상태와 특화 shadow 경로로 줄여라.
Applicability: 클라우드 네이티브 secure container.
Limits: x86/KVM 및 게스트 수정, shadow coherence 구현에 의존하고 공개 정량치가 제한적이다. +
Verification: abstract_only; confidence=high.
Canonical title: PVM: Efficient Shadow Paging for Deploying Secure Containers in Cloud-native Environment
Question: 하드웨어 가상화 기능을 게스트에 노출하지 않고 안전 컨테이너의 메모리 가상화를 빠르게 할 수 있는가?
Context: nested virtualization은 복잡·비싸고 전통 shadow paging은 동시성·일관성 병목이 있다.
Method: PVM은 게스트 하이퍼바이저와 최소 공유 영역, 효율적 shadow paging을 사용한다.
Evaluation: workloads=Alibaba/Ant production, tens of thousands containers/day; CPU, memory, I/O benchmarks; baselines=nested KVM; metrics=memory virtualization performance; CPU/I/O performance; results=Outperforms nested KVM for concurrent memory virtualization; exact aggregate not abstract-verified.
Interpretation: 보안 경계를 유지하면서 게스트에 필요한 최소 협업 인터페이스만 노출할 수 있다.
Reusable lesson: nested 계층은 최소 공유 상태와 특화 shadow 경로로 줄여라.
Applicability: 클라우드 네이티브 secure container.
Limits: x86/KVM 및 게스트 수정, shadow coherence 구현에 의존하고 공개 정량치가 제한적이다. +
Bibliographic paper record. +
r
Verification: official USENIX paper page and abstract; confidence=high.
Canonical title: Rearchitecting Buffered I/O in the Era of High-Bandwidth SSDs
Question: How can buffered writes exploit high-bandwidth SSDs without losing the usability and read benefits of the page cache?
Context: Caching every write, serialized page management, and read-before-write for partial pages limit modern SSD bandwidth.
Method: WSBuffer buffers only small/unaligned write scraps, sends aligned large regions directly, and uses two-stage flushing plus concurrent page management.
Evaluation: workloads=buffered-I/O workloads on high-bandwidth SSDs; baselines=ext4, F2FS, Btrfs, XFS, ScaleCache; metrics=throughput and latency; results=up to 3.91x throughput and 82.80x latency improvement
Interpretation: Buffered I/O need not force all bytes through page-sized caching on the write path.
Reusable lesson: Buffer only device-unfriendly fragments and direct-transfer naturally aligned bulk data.
Applicability: Linux filesystems on high-bandwidth SSDs, especially partial/unaligned writes.
Limits: Headline results focus on evaluated filesystems/devices; adds a new buffering/flushing architecture. +
Bibliographic paper record. +
Verification: official USENIX paper page and abstract; confidence=medium.
Canonical title: Rearchitecting Buffered I/O in the Era of High-Bandwidth SSDs
Question: How can buffered writes exploit high-bandwidth SSDs without losing the usability and read benefits of the page cache?
Context: Caching every write, serialized page management, and read-before-write for partial pages limit modern SSD bandwidth.
Method: WSBuffer buffers only small/unaligned write scraps, sends aligned large regions directly, and uses two-stage flushing plus concurrent page management.
Evaluation: workloads=buffered-I/O workloads on high-bandwidth SSDs; baselines=ext4, F2FS, Btrfs, XFS, ScaleCache; metrics=throughput and latency; results=up to 3.91x throughput and 82.80x latency improvement
Interpretation: Buffered I/O need not force all bytes through page-sized caching on the write path.
Reusable lesson: Buffer only device-unfriendly fragments and direct-transfer naturally aligned bulk data.
Applicability: Linux filesystems on high-bandwidth SSDs, especially partial/unaligned writes.
Limits: Headline results focus on evaluated filesystems/devices; adds a new buffering/flushing architecture. +
정본 Lesson 보강 근거. 검토 원본: Lesson:technical_review_request_aware_cooperative_i_o_scheduling_for_scale_out_database_applications_fe4a3e09. 확인 범위: full_text. 확인한 자료: https://www.usenix.org/conference/hotstorage17/program/presentation/jo ; https://www.usenix.org/system/files/conference/hotstorage17/hotstorage17-paper-jo.pdf. 질문, 방법, 평가, 해석, 재사용 교훈, 적용 범위와 한계를 같은 Lesson 본문에 통합했습니다. +