속성:Evidence note
자료에서 확인할 수 있는 점이나 한계를 적습니다.
a
정본 Lesson 보강 근거. 검토 원본: Lesson:technical_review_analysis_of_virtual_machine_live_migration_as_a_method_for_power_capping_b973b1aa. 확인 범위: official_abstract. 확인한 자료: https://doi.org/10.1007/s11227-013-0956-1 ; https://yonsei.elsevierpure.com/en/publications/analysis-of-virtual-machine-live-migration-as-a-method-for-power-/ ; https://koasas.kaist.ac.kr/handle/10203/188588. 질문, 방법, 평가, 해석, 재사용 교훈, 적용 범위와 한계를 같은 Lesson 본문에 통합했습니다. +
Source: Yonsei University Computer Systems Laboratory publication list supplied by the user. Manifestation 1 of 1. +
Verification: abstract_only; confidence=high.
Canonical title: Antipode: Enforcing Cross-Service Causal Consistency in Distributed Applications
Question: 여러 서비스·데이터 저장소를 가로지르는 인과 일관성을 애플리케이션에 투명하게 보장할 수 있는가?
Context: 서비스별 저장소는 로컬 일관성만 제공해 교차 서비스 순서 위반이 생긴다.
Method: Antipode는 요청과 저장소 내부로 연산 lineage를 전파하고 라이브러리로 점진 도입한다.
Evaluation: workloads=8 datastores; 2 microservice benchmarks; baselines=uncoordinated services; metrics=inconsistencies; latency; throughput; results=Prevents cross-service inconsistencies with <2% performance impact.
Interpretation: 의존성 메타데이터를 종단간 전파하면 서비스 경계를 넘어 인과성을 보존할 수 있다.
Reusable lesson: 교차 서비스 일관성은 각 계층의 lineage 전달 계약이 핵심이다.
Applicability: 마이크로서비스와 이종 데이터스토어.
Limits: 라이브러리·저장소 통합이 필요하고 인과 일관성보다 강한 의미론은 다루지 않는다. +
Verification: abstract_only; confidence=medium.
Canonical title: Antipode: Enforcing Cross-Service Causal Consistency in Distributed Applications
Question: 여러 서비스·데이터 저장소를 가로지르는 인과 일관성을 애플리케이션에 투명하게 보장할 수 있는가?
Context: 서비스별 저장소는 로컬 일관성만 제공해 교차 서비스 순서 위반이 생긴다.
Method: Antipode는 요청과 저장소 내부로 연산 lineage를 전파하고 라이브러리로 점진 도입한다.
Evaluation: workloads=8 datastores; 2 microservice benchmarks; baselines=uncoordinated services; metrics=inconsistencies; latency; throughput; results=Prevents cross-service inconsistencies with <2% performance impact.
Interpretation: 의존성 메타데이터를 종단간 전파하면 서비스 경계를 넘어 인과성을 보존할 수 있다.
Reusable lesson: 교차 서비스 일관성은 각 계층의 lineage 전달 계약이 핵심이다.
Applicability: 마이크로서비스와 이종 데이터스토어.
Limits: 라이브러리·저장소 통합이 필요하고 인과 일관성보다 강한 의미론은 다루지 않는다. +
Bibliographic paper record. +
Source: Yonsei University Computer Systems Laboratory publication list supplied by the user. Manifestation 1 of 2. +
Source: Yonsei University Computer Systems Laboratory publication list supplied by the user. Manifestation 2 of 2. +
정본 Lesson 보강 근거. 검토 원본: Lesson:technical_review_application_aware_swapping_for_mobile_systems_3dd6bfc2. 확인 범위: official_abstract. 확인한 자료: https://doi.org/10.1145/3126509 ; https://www.sigbed.org/emsoft-info/confs/emsoft2017.html ; https://yonsei.elsevierpure.com/en/publications/application-aware-swapping-for-mobile-systems/. 질문, 방법, 평가, 해석, 재사용 교훈, 적용 범위와 한계를 같은 Lesson 본문에 통합했습니다. +
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_appwatch_detecting_kernel_bug_for_protecting_consumer_electronics_applications_779775f3. 확인 범위: official_abstract. 확인한 자료: https://doi.org/10.1109/TCE.2010.5505989 ; https://yonsei.elsevierpure.com/en/publications/appwatch-detecting-kernel-bug-for-protecting-consumer-electronics/. 질문, 방법, 평가, 해석, 재사용 교훈, 적용 범위와 한계를 같은 Lesson 본문에 통합했습니다. +
archer adaptive memory compression with page association rule awareness for high speed response 98bcc760 +
Bibliographic paper record. +
archer adaptive memory compression with page association rule awareness for high speed response 98bcc760 +
Verification: official USENIX page and abstract; confidence=high.
Canonical title: Archer: Exploiting Memory Content Correlation for Mobile Memory Compression
Question: Can correlations among anonymous pages improve mobile compressed-memory capacity and responsiveness?
Context: Conventional per-page compression misses repeated/correlated content and forces a fixed granularity.
Method: Archer discovers association rules, selects flexible compression granularity, and redesigns LRU management with an adaptive compression region.
Evaluation: workloads=mobile application launch, photo processing, and frame-rate workloads; baselines=state-of-the-art mobile memory compression; metrics=correlated-page share and application performance; results=about 25% correlated; 1.55x, 1.42x, and 1.31x improvements
Interpretation: Cross-page semantic/content relationships can be converted into practical compressed-memory gains.
Reusable lesson: Choose compression units from observed redundancy rather than VM page boundaries alone.
Applicability: Mobile anonymous memory with recurring or correlated page content.
Limits: Correlation discovery and benefits depend on app mix; results target mobile workloads. +
archer adaptive memory compression with page association rule awareness for high speed response 98bcc760 +
Verification: official USENIX page and abstract; confidence=medium.
Canonical title: Archer: Adaptive Memory Compression with Page-Association-Rule Awareness for High-Speed Response of Mobile Devices
Question: Can correlations among anonymous pages improve mobile compressed-memory capacity and responsiveness?
Context: Conventional per-page compression misses repeated/correlated content and forces a fixed granularity.
Method: Archer discovers association rules, selects flexible compression granularity, and redesigns LRU management with an adaptive compression region.
Evaluation: workloads=mobile application launch, photo processing, and frame-rate workloads; baselines=state-of-the-art mobile memory compression; metrics=correlated-page share and application performance; results=about 25% correlated; 1.55x, 1.42x, and 1.31x improvements
Interpretation: Cross-page semantic/content relationships can be converted into practical compressed-memory gains.
Reusable lesson: Choose compression units from observed redundancy rather than VM page boundaries alone.
Applicability: Mobile anonymous memory with recurring or correlated page content.
Limits: Correlation discovery and benefits depend on app mix; results target mobile workloads. +
정본 Lesson 보강 근거. 검토 원본: Lesson:technical_review_asap_fast_mobile_application_switch_via_adaptive_prepaging_3ea3e879. 확인 범위: full_text. 확인한 자료: https://www.usenix.org/conference/atc21/presentation/son ; https://www.usenix.org/system/files/atc21-son.pdf. 질문, 방법, 평가, 해석, 재사용 교훈, 적용 범위와 한계를 같은 Lesson 본문에 통합했습니다. +
Source: Yonsei University Computer Systems Laboratory publication list supplied by the user. Manifestation 1 of 1. +
Verification: official_abstract; confidence=medium.
Canonical title: Asymmetric RAID: Rethinking RAID for SSD Heterogeneity
Question: Can RAID use heterogeneous SSD capacity and performance without forcing symmetric participation in every stripe?
Context: Conventional RAID assumes similar devices and lets the weakest member constrain capacity or speed.
Method: Asymmetric RAID distributes data unevenly and exposes different portions of stripe address space to devices for placement according to their characteristics.
Evaluation: workloads=; baselines=symmetric RAID; metrics=capacity utilization; performance; rebuild behavior; results=no quantitative result verified
Interpretation: Stripe geometry can be a placement policy rather than a fixed equal partition.
Reusable lesson: Represent heterogeneous capacity/performance explicitly instead of normalizing every device to the minimum.
Applicability: RAID arrays containing heterogeneous SSDs.
Limits: Preliminary workshop work; accessible primary sources do not provide a complete evaluation, recovery analysis, or numbers. +
Verification: official_abstract; confidence=medium.
Canonical title: Asymmetric RAID: Rethinking RAID for SSD Heterogeneity
Question: Can RAID use heterogeneous SSD capacity and performance without forcing symmetric participation in every stripe?
Context: Conventional RAID assumes similar devices and lets the weakest member constrain capacity or speed.
Method: Asymmetric RAID distributes data unevenly and exposes different portions of stripe address space to devices for placement according to their characteristics.
Evaluation: workloads=; baselines=symmetric RAID; metrics=capacity utilization; performance; rebuild behavior; results=no quantitative result verified
Interpretation: Stripe geometry can be a placement policy rather than a fixed equal partition.
Reusable lesson: Represent heterogeneous capacity/performance explicitly instead of normalizing every device to the minimum.
Applicability: RAID arrays containing heterogeneous SSDs.
Limits: Preliminary workshop work; accessible primary sources do not provide a complete evaluation, recovery analysis, or numbers. +
Bibliographic paper record. +
정본 Lesson 보강 근거. 검토 원본: Lesson:technical_review_asynchronous_i_o_stack_a_low_latency_kernel_i_o_stack_for_ultra_low_latency_ssd_3852c43e. 확인 범위: full_text. 확인한 자료: https://www.usenix.org/conference/atc19/presentation/lee-gyusun ; https://www.usenix.org/system/files/atc19-lee-gyusun.pdf. 질문, 방법, 평가, 해석, 재사용 교훈, 적용 범위와 한계를 같은 Lesson 본문에 통합했습니다. +
Verification: official USENIX page and abstract; confidence=high.
Canonical title: Asynchronous I/O Stack: A Low-latency Kernel I/O Stack for Ultra-Low Latency SSDs
Question: How should the Linux kernel I/O path change when software overhead rivals ultra-low-latency SSD access time?
Context: Synchronous layers, heavyweight block processing, and page-cache handling mask the latency of fast NVMe/Optane devices.
Method: AIOS redesigns the Linux path around asynchronous processing and lightweight NVMe block-layer and page-cache mechanisms.
Evaluation: workloads=FIO, RocksDB, Filebench, and OLTP on Optane; baselines=standard Linux I/O stack; metrics=latency, IOPS, throughput; results=15–33% lower latency; 11–44% more RocksDB IOPS; 15–30% more application throughput
Interpretation: Faster media shifts the bottleneck into kernel control flow and data-structure overhead.
Reusable lesson: Remove serialization and general-purpose layering once device latency approaches software-path cost.
Applicability: Linux storage stacks for ultra-low-latency NVMe/Optane-class media.
Limits: Requires kernel changes; reported gains reflect specific devices and workloads. +