{LegoOS}: A disseminated, distributed {OS} for hardware resource disaggregation

Y Shan, Y Huang, Y Chen, Y Zhang - 13th USENIX Symposium on …, 2018 - usenix.org
The monolithic server model where a server is the unit of deployment, operation, and failure
is meeting its limits in the face of several recent hardware and application trends. To improve …

Semeru: A {Memory-Disaggregated} managed runtime

C Wang, H Ma, S Liu, Y Li, Z Ruan, K Nguyen… - … USENIX Symposium on …, 2020 - usenix.org
Resource-disaggregated architectures have risen in popularity for large datacenters.
However, prior disaggregation systems are designed for native applications; in addition, all …

Kvell: the design and implementation of a fast persistent key-value store

B Lepers, O Balmau, K Gupta… - Proceedings of the 27th …, 2019 - dl.acm.org
Modern block-addressable NVMe SSDs provide much higher bandwidth and similar
performance for random and sequential access. Persistent key-value stores (KVs) designed …

ThyNVM: Enabling software-transparent crash consistency in persistent memory systems

J Ren, J Zhao, S Khan, J Choi, Y Wu… - Proceedings of the 48th …, 2015 - dl.acm.org
Emerging byte-addressable nonvolatile memories (NVMs) promise persistent memory,
which allows processors to directly access persistent data in main memory. Yet, persistent …

Assise: Performance and availability via client-local {NVM} in a distributed file system

TE Anderson, M Canini, J Kim, D Kostić… - … USENIX Symposium on …, 2020 - usenix.org
The adoption of low latency persistent memory modules (PMMs) upends the long-
established model of remote storage for distributed file systems. Instead, by colocating …

Flash storage disaggregation

A Klimovic, C Kozyrakis, E Thereska, B John… - Proceedings of the …, 2016 - dl.acm.org
PCIe-based Flash is commonly deployed to provide datacenter applications with high IO
rates. However, its capacity and bandwidth are often underutilized as it is difficult to design …

I {Can't} Believe {It's} Not Causal! Scalable Causal Consistency with No Slowdown Cascades

SA Mehdi, C Littley, N Crooks, L Alvisi… - … USENIX Symposium on …, 2017 - usenix.org
We describe the design, implementation, and evaluation of Occult (Observable Causal
Consistency Using Lossy Timestamps), the first scalable, geo-replicated data store that …

An in-depth analysis of cloud block storage workloads in large-scale production

J Li, Q Wang, PPC Lee, C Shi - 2020 IEEE International …, 2020 - ieeexplore.ieee.org
Cloud block storage systems support diverse types of applications in modern cloud services.
Characterizing their I/O activities is critical for guiding better system designs and …

Verifying a high-performance crash-safe file system using a tree specification

H Chen, T Chajed, A Konradi, S Wang, A İleri… - Proceedings of the 26th …, 2017 - dl.acm.org
DFSCQ is the first file system that (1) provides a precise specification for fsync and
fdatasync, which allow applications to achieve high performance and crash safety, and (2) …

Seeing is believing: A client-centric specification of database isolation

N Crooks, Y Pu, L Alvisi, A Clement - Proceedings of the ACM …, 2017 - dl.acm.org
This paper introduces the first state-based formalization of isolation guarantees. Our
approach is premised on a simple observation: applications view storage systems as black …