{ZNS}: Avoiding the block interface tax for flash-based {SSDs}
M Bjørling, A Aghayev, H Holmberg… - 2021 USENIX Annual …, 2021 - usenix.org
The Zoned Namespace (ZNS) interface represents a new division of functionality between
host software and flash-based SSDs. Current flash-based SSDs maintain the decades-old …
host software and flash-based SSDs. Current flash-based SSDs maintain the decades-old …
{LightNVM}: The linux {Open-Channel}{SSD} subsystem
As Solid-State Drives (SSDs) become commonplace in data-centers and storage arrays,
there is a growing demand for predictable latency. Traditional SSDs, serving block I/Os, fail …
there is a growing demand for predictable latency. Traditional SSDs, serving block I/Os, fail …
Why does the cloud stop computing? lessons from hundreds of service outages
We conducted a cloud outage study (COS) of 32 popular Internet services. We analyzed
1247 headline news and public post-mortem reports that detail 597 unplanned outages that …
1247 headline news and public post-mortem reports that detail 597 unplanned outages that …
Tiny-tail flash: Near-perfect elimination of garbage collection tail latencies in NAND SSDs
Flash storage has become the mainstream destination for storage users. However, SSDs do
not always deliver the performance that users expect. The core culprit of flash performance …
not always deliver the performance that users expect. The core culprit of flash performance …
Making disk failure predictions {SMARTer}!
Disk drives are one of the most commonly replaced hardware components and continue to
pose challenges for accurate failure prediction. In this work, we present analysis and …
pose challenges for accurate failure prediction. In this work, we present analysis and …
File systems unfit as distributed storage backends: lessons from 10 years of Ceph evolution
For a decade, the Ceph distributed file system followed the conventional wisdom of building
its storage backend on top of local file systems. This is a preferred choice for most distributed …
its storage backend on top of local file systems. This is a preferred choice for most distributed …
Fail-slow at scale: Evidence of hardware performance faults in large production systems
HS Gunawi, RO Suminto, R Sears, C Golliher… - ACM Transactions on …, 2018 - dl.acm.org
Fail-slow hardware is an under-studied failure mode. We present a study of 114 reports of
fail-slow hardware incidents, collected from large-scale cluster deployments in 14 …
fail-slow hardware incidents, collected from large-scale cluster deployments in 14 …
The {CASE} of {FEMU}: Cheap, accurate, scalable and extensible flash emulator
We present FEMU, a QEMU-based flash emulator for fostering future full-stack
software/hardware SSD research, with the following four" CASE" benefits. FEMU is cheap …
software/hardware SSD research, with the following four" CASE" benefits. FEMU is cheap …
TaxDC: A taxonomy of non-deterministic concurrency bugs in datacenter distributed systems
We present TaxDC, the largest and most comprehensive taxonomy of non-deterministic
concurrency bugs in distributed systems. We study 104 distributed concurrency (DC) bugs …
concurrency bugs in distributed systems. We study 104 distributed concurrency (DC) bugs …
{LinnOS}: Predictability on unpredictable flash storage with a light neural network
This paper presents LinnOS, an operating system that leverages a light neural network for
inferring SSD performance at a very fine—per-IO—granularity and helps parallel storage …
inferring SSD performance at a very fine—per-IO—granularity and helps parallel storage …