Differentiated storage services

Mesnier, Michael P.; Akers, Jason

doi:10.1145/1945023.1945030

Cited by 41 publications

(34 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Multi-stream SSD [16] reduces GC overhead by maintaining multiple streams in an SSD according to the expected lifetime of the data. A similar study has also been conducted to provide differentiated services by sending requests from hosts to storage devices along with classification tags [4]. In contrast, we focus analyzing the effect of semantic information of flash storage devices in terms of performance and lifetime in this paper.…”

Section: Background and Related Workmentioning

confidence: 99%

“…Manuscript received March 7, 2017 Recently, SCSI, eMMC 4.5, and Universal Flash Storage (UFS 2.0) standards provide semantic information such as SCSI Group Number, Stream ID, Context ID and Data Tag to the interface between a host system and a storage device [4]- [7]. We can expect to improve performance of flash storage devices by using such information.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Analysis of Performance for NAND Flash Based SSDs via Using Host Semantic Information

Kim

Park

2017

IEICE Trans. Inf. & Syst.

View full text Add to dashboard Cite

SUMMARYThe use of flash memory based storage devices is rapidly increasing, and user demands for high performance are also constantly increasing. The performance of the flash storage device is greatly influenced by cleaning operations of Flash Translation Layer (FTL). Various studies have been conducted to lower the cost of cleaning operations. However, there are limits to achieve sufficient performance improvement of flash storages without help of a host system, with only limited information in storage devices. Recently, SCSI, eMMC, and UFS standards provide an interface for sending semantic information from a host system to a storage device.In this paper, we analyze effects of semantic information on performance and lifetime of flash storage devices. We evaluate performance and lifetime improvement through SA-FTL (Semantic Aware Flash Translation Layer), which can take advantage of semantic information in storage devices. Experiments show that SA-FTL improves performance and lifetime of flash based storages by up to 30 and 35%, respectively, compared to a simple page-level FTL.

show abstract

Section: Background and Related Workmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Analysis of Performance for NAND Flash Based SSDs via Using Host Semantic Information

Kim

Park

2017

IEICE Trans. Inf. & Syst.

View full text Add to dashboard Cite

show abstract

“…With a set of predefined rules, each type is associated with a QoS policy that can be supported by the underlying storage system. At runtime, using the Differentiated Storage Services [18] protocol, the associated policy of a request is delivered to the storage system along with the request itself. Upon receiving a request, the storage system first extracts the associated QoS policy, and then uses a proper mechanism to serve the request as required by the QoS policy.…”

Section: Our Solutionmentioning

confidence: 99%

“…The QoS policy is embedded into the original I/O request and delivered to the storage system through a block interface. We have implemented Hetero-DB by using the Differentiated Storage Services protocol from Intel Labs [18] to deliver a request and its associated policy to a hybrid storage system. Upon receiving a request, the storage system first extracts the policy, and invokes a mechanism to serve this request.…”

Section: Gpu-aware Query Execution Enginementioning

confidence: 99%

Hetero-DB: Next Generation High-Performance Database Systems by Best Utilizing Heterogeneous Computing and Storage Resources

Zhang

Chen

Ding

et al. 2015

J. Comput. Sci. Technol.

View full text Add to dashboard Cite

With recent advancement on hardware technologies, new general-purpose high-performance devices have been widely adopted, such as the graphics processing unit (GPU) and solid state drive (SSD). GPU may offer an order of higher throughput for applications with massive data parallelism, compared with the multicore CPU. Moreover, new storage device SSD is also capable of offering a much higher I/O throughput and lower latency than a traditional hard disk device (HDD). These new hardware devices can significantly boost the performance of many applications; thus the database community has been actively engaging in adopting them into database systems. However, the performance benefit cannot be easily reaped if the new hardwares are improperly used. In this paper, we propose Hetero-DB, a high-performance database system by exploiting both the characteristics of the database system and the special properties of the new hardware devices in system's design and implementation. Hetero-DB develops a GPU-aware query execution engine with GPU device memory management and query scheduling mechanism to support concurrent query execution. Furthermore, with the SSD-HDD hybrid storage system, we redesign the storage engine by organizing HDD and SSD into a two-level caching hierarchy in Hetero-DB. To best utilize the hybrid hardware devices, the semantic information that is critical for storage I/O is identified and passed to the storage manager, which has a great potential to improve the efficiency and performance. Hetero-DB aims to maximize the performance benefits of GPU and SSD, and demonstrates the effectiveness for designing next generation database systems.

show abstract

“…These systems provide latency guarantees, but not for 99 th percentile latency. There are also many other systems which focus on fair sharing and quality-of-service in the storage context [10,33,27,20,9,30].…”

Section: Related Workmentioning

confidence: 99%

Cake: Enabling High-level SLOs on Shared Storage Systems

Wang¹,

Venkataraman²,

Alspaugh³

et al. 2012

View full text Add to dashboard Cite

Cake is a coordinated, multi-resource scheduler for shared distributed storage environments with the goal of achieving both high throughput and bounded latency. Cake uses a two-level scheduling scheme to enforce high-level service-level objectives (SLOs). Firstlevel schedulers control consumption of resources such as disk and CPU. These schedulers (1) provide mechanisms for differentiated scheduling, (2) split large requests into smaller chunks, and (3) limit the number of outstanding device requests, which together allow for effective control over multi-resource consumption within the storage system. Cake's second-level scheduler coordinates the first-level schedulers to map high-level SLO requirements into actual scheduling parameters. These parameters are dynamically adjusted over time to enforce high-level performance specifications for changing workloads. We evaluate Cake using multiple workloads derived from real-world traces. Our results show that Cake allows application programmers to explore the latency vs. throughput trade-off by setting different high-level performance requirements on their workloads. Furthermore, we show that using Cake has concrete economic and business advantages, reducing provisioning costs by up to 50% for a consolidated workload and reducing the completion time of an analytics cycle by up to 40%.

show abstract

Differentiated storage services

Cited by 41 publications

References 22 publications

Analysis of Performance for NAND Flash Based SSDs via Using Host Semantic Information

Analysis of Performance for NAND Flash Based SSDs via Using Host Semantic Information

Hetero-DB: Next Generation High-Performance Database Systems by Best Utilizing Heterogeneous Computing and Storage Resources

Cake: Enabling High-level SLOs on Shared Storage Systems

Contact Info

Product

Resources

About