Figure 4 - uploaded by Scott A. Brandt
Content may be subject to copyright.
Source publication
We can outperform the most powerful disk spin-down algorithm available by using a small non-volatile storage write-only cache to temporarily store hard disk write re-quests while a disk is spun-down. In our experiments, we use a USB thumb drive device as our non-volatile storage cache. By redirecting writes to a flash memory while a disk is spun-do...
Contexts in source publication
Context 1
... experiment uses a 64MB USB thumb drive from Dell. Figure 4 shows the results of our experiment. In this figure, Read USB is the time it takes to read the data blocks from the USB storage device into memory. ...
Context 2
... cwr at 25, we can save up to 36% energy, and with cwr at 150, we can save up to 46% en- ergy. The cwr value shouldn't just be set at 150 because as we mentioned before, the more data we redirect the longer it takes to synchronize the disk with the USB data blocks as we saw in figure 4. ...
Citations
... Moreover, in perspective of managing and reducing power consumption, Kim et al. [12], by placing specific information on flash chips instead of the traditional hard disks, and also Bisson [13] and Chen [14], by using USB flash disks as the cache of traditional hard disks, were successful. ...
Storing digital information, ensuring the accuracy, steady and uninterrupted
access to the data are considered as fundamental challenges in enterprise-class
organizations and companies. In recent years, new types of storage systems such
as solid state disks (SSD) have been introduced. Unlike hard disks that have
mechanical structure, SSDs are based on flash memory and thus have electronic
structure. Generally a SSD consists of a number of flash memory chips, some
buffers of the volatile memory type, and an embedded microprocessor, which have
been interconnected by a port. This microprocessor run a small file system
which called flash translation layer (FTL). This software controls and
schedules buffers, data transfers and all flash memory tasks. SSDs have some
advantages over hard disks such as high speed, low energy consumption, lower
heat and noise, resistance against damage, and smaller size. Besides, some
disadvantages such as limited endurance and high price are still challenging.
In this study, the effort is to combine two common technologies - SLC and MLC
chips - used in the manufacture of SSDs in a single SSD to decrease the side
effects of current SSDs. The idea of using multi-layer SSD is regarded as an
efficient solution in this field.
... The trade-offs associated with hard disks and flash memory devices, either in the form of price, capacity, or performance, motivate lots of storage system designs. Many researchers have proposed the use of flash memory as a non-volatile cache to prolong the disk spin down time[3] [4] [5] [6] [7] [8]. However, these schemes treat flash memory as complement of DRAM buffer cache, and only a subset of data blocks are cached in flash memory, then the disk is used quite frequently due to cache misses or flushing. ...
... The different features of flash memory and hard disk have motivated lots of researches. A typical method is using flash memory as non-volatile secondary buffer cache which maintains blocks to be accessed in the near future[3] [4] [5] [6] [7] [8]. March et al.[3] place the flash memory between DRAM and disk. ...
... All read and write requests to disk are cached in the flash memory. Bission et al.[4] [5] focus on the redirection of write requests to a flash memory device when the hard disk is spun-down. Chen et al.[6] partition the flash memory into a read cache, a prefetch buffer, and a write buffer to save energy. ...
Flash memory is widely used because of its shock-resistance and power-efficient features. However, it cannot replace hard disks as secondary storage devices due to their greater cost per unit storage and low capability. In this paper, we propose an energy efficient heterogeneous secondary storage system management scheme for mobile systems. We employ flash memory device as a file cache of hard disk and extend existing data cache management algorithms to distribute files between two devices with consideration of file level cache restrictions. As a result, most file accesses are conducted in flash memory device and disk is spun-down to save energy. We develop a trace-driven simulator to evaluate our scheme in comparison with other alternatives. Results demonstrate that with the help of our scheme, energy consumption of secondary storage system can be saved by up to 90% and I/O access time is improved. Furthermore, the file cache management algorithms can result in high hit ratios.
... In [12,2,3,4] different techniques are proposed to use Flash memory as non-volatile cache. The aim is to maintain blocks which are likely to be accessed in the near future in Flash memory. ...
We propose a new type of heterogeneous storage device called Combo Drive, which comprises of a smaller-capacity low-latency solid-state disk drive (SSD) concatenated with a larger-capacity high-throughput hard disk drive (HDD). The overall cost of a Combo Drive, similar to a Hybrid Drive, is still dominated by the more capacious HDD. With Combo Drive, the performance advantages of both the SSD and the HDD are readily utilized by assigning the lower portion of the address space, which is already considered by many file systems as faster than the higher portion, to the SSD. Performance can be optimized further on file system level on the host side. In contrast, existing Hybrid Drives utilize non-volatile memory hierarchically as a cache transparent to the environment requiring complex cache coherence al-gorithms. We built a Combo Drive prototype and propose multiple heuristic optimization algorithms implemented in file-system-level optimizers. Performance measurements on the host side show that the prototype achieves system start up time and application launch time similar to an SSD alone while offering large capacity and low cost of an HDD.
... Bisson et al , in [6] and [5], reduce the energy consumption and increase the efficiency of hard drive spin down algorithms through the addition of a flash memory cache. Although they do make use of flash memories, it is used as a SRAM R 1,1 R 1,2 R 1,3 R 1,… R 2,1 R 2,2 R 2,3 R 2,… R …,1 R …,2 R …,3 R …,… R n,1 R n,2 R n, 3 (a) The sensor system: each process generates a series of requests, which are placed into the scheduler residing in SRAM. ...
Flash equipped sensor devices are becoming increasingly complex and are now capable of supporting real-time multiple applications
on a single sensor, rich sensing of visual and audio data, and storage of large amounts of data. With this increase in complexity,
it is no longer sufficient to provide first in first out (FIFO) type capture of data into more persistent memories. In this
paper we propose RG-EDF, a new scheduling policy for flash equipped sensor devices. RG-EDF aims at providing QoS support to
multimedia tasks by considering the unique characteristics of flash-based devices. We have implemented our scheme on a CC1010
sensor node with a SD flash card attached and compared our technique to other popular scheduling policies. Our experimental
results show the working and benefit of our system.
... In [7], Singleton, Nathuji, and Schwan explore algorithms to manage flash allocation as a cache for disk access. Similarly, Bisson and Brandt in [3] examined similar aspects to allow for improved disk spin down for reduced energy consumption. Chen, Jiang, and Zhang propose SmartSaver in [4] for using flash as a pre- fetch and cache for web content for mobile nodes. ...
Even though their capacities are still orders of magnitude lower than those of hard disks, flash storage systems are rapidly gaining importance in energy-constrained systems. This paper focuses on USB flash drives, which can provide portable storage to mobile systems or storage to systems that otherwise do not have persistent storage opportunities (e.g., low-power sensor devices). The paper presents studies relating to power consumption, energy overheads and benefits, and performance impacts of USB flash drives. The key insights obtained from these experiments are that (i) read/write costs are not significantly greater than idle costs and (ii) the size of the flash itself has only limited bearing on energy consumption.
... In all experiments, EXCES was configured to use an ECD miss rate threshold of 1000 misses-perminute to trigger reconfiguration and a minimum duration of one minute between two reconfiguration operations. 3 We used the BLTK (Linux Battery Life Tool Kit) [5] as our primary benchmark for system evaluation. This benchmark focuses specifically on laptop-specific workloads, targeted for evaluating battery life of laptop systems in realistic usage scenarios. ...
... Figure 8 shows the results using timeout intervals of 5, 10, 15, 30 seconds and no timeout (∞ seconds). We used 3 While we used these static values (based on preliminary experimentation) for simplicity, subsequent versions of EXCES will be able to dynamically adapt these thresholds based on application workload. hdparm to set the timeout intervals in the disk's firmware, restricted to a minimum value of 5 seconds. ...
... Further, our findings point to a range of ∼2-14% for the cases when EXCES was indeed able to reduce power consumption, in somewhat of a contrast to earlier results from simulation studies [3,8,18] that predicted energy savings of ∼20-46% 4 . This difference is primarily because the powerconsumption of the ECD was considered negligible and ignored in those studies. ...
Power consumption within the disk-based storage subsystem forms a substantial portion of the overall energy footprint in commodity systems. Researchers have proposed external caching on a persistent, low-power storage device, which we term external caching device (ECD), to minimize disk activity and conserve energy. While recent simulation-based studies have argued in favor of this approach, the lack of an actual system implementation has precluded answering several key questions about external caching systems. We present the design and implementation of EXCES, an external caching system that employs prefetching, caching, and buffering of disk data for reducing disk activity. EXCES addresses important questions related to external caching, including the estimation of future data popularity, I/O indirection, continuous reconfiguration of the ECD contents, and data consistency. We evaluated EXCES with both micro- and macro-benchmarks that address idle, I/O intensive, and real-world workloads. Overall system energy savings was found to lie in the modest 2-14% range, depending on the workload, in somewhat of a contrast to the higher values predicted by earlier studies. Furthermore, while the CPU and memory overheads of EXCES were well within acceptable limits, we found that flash-based external caching can substantially degrade I/O performance. We believe that external caching systems hold promise. Further improvements in ECD technology, both in terms of their power consumption and performance characteristics can help realize the full potential of such systems.
... I/O performance of the storage subsystem with a higher energy efficiency. These proposals can be grouped in two categories: 1) ones using a flash as a non-volatile cache (NVC) [3, 4, 5, 6], which stores data blocks which are likely to be accessed in the near future, allowing the disk to spin down for longer periods, and 2) the others using flash as a component of heterogeneous secondary storage with data concentration techniques which minimize slower disk traffics [7]. A hybrid hard disk [8, 9, 10], which uses a flash memory as an onboard NVC, is the most well-known example of the first category above. ...
... There has been a lot of research on combining hard disks with flash memory in mobile storage systems. March et al. [3], Bisson et al. [4], Chen et al. [5], and Kgil et al. [6], have all proposed using flash memory as an NVC, to store blocks which are likely to be accessed in the near future, and thus allowing a hard disk to spin down for longer. Bisson et al. [4] focused on the redirection of write requests to a flash memory instead of a hard disk, while Chen et al. recently studied partitioning a flash memory into a cache, a prefetch buffer, and a write buffer to save energy. ...
... March et al. [3], Bisson et al. [4], Chen et al. [5], and Kgil et al. [6], have all proposed using flash memory as an NVC, to store blocks which are likely to be accessed in the near future, and thus allowing a hard disk to spin down for longer. Bisson et al. [4] focused on the redirection of write requests to a flash memory instead of a hard disk, while Chen et al. recently studied partitioning a flash memory into a cache, a prefetch buffer, and a write buffer to save energy. Kgil et al. focused on reducing main memory power consumption by using a flash memory as a second-level buffer cache. ...
A hybrid hard disk employs the advantages of both a hard disk and a NAND flash memory, thus making it a cost-effective fast secondary storage device. In this paper, we improve its I/O performance by combining an intelligent data pinning policy for the flash memory with a caching technique which is aware of access patterns for the flash memory and DRAM. Our proposed techniques reduce the system boot time and application launching time while reducing energy consumption, which is vital in the mobile devices. We built SimHybrid, a flexible trace-driven hybrid hard disk evaluation environment, and used it to demonstrate how a hybrid hard disk can achieve significantly better I/O performance than a traditional hard disk while using much less energy.
... If all the data have to go through the flash memory, its low write bandwidth and limited erasure cycles would pose a serious problem. Another work [1] proposes to redirect write-back traffic to the flash drive when the disk is spun down so that the number of costly disk spin-up/downs could be reduced. However, this approach simply uses the flash drive as a writeonly cache and it has no caching or prefetching mechanisms. ...
In a mobile computer the hard disk consumes a considerable amount of energy. Existing dynamic power management policies usually take conservative approaches to save disk energy, and disk energy consumption remains a serious issue. Meanwhile, the flash drive is becoming a must-have portable storage device for almost every laptop user on travel. In this paper, we propose to make another highly desired use of the flash drive saving disk energy. This is achieved by using the flash drive as a standby buffer for caching and prefetching disk data. Our design significantly extends disk idle times with careful and deliberate consideration of the particular characteristics of the flash drive. Trace-driven simulations show that up to 41% of disk energy can be saved with a relatively small amount of data written to the flash drive.
... The different characteristics of flash memory and HDDs have motivated a lot of research on energy-efficient mobile storage systems which combine disks and flash memory. Many researchers have proposed the use of flash memory as a nonvolatile cache [4, 5, 6, 7, 2, 8], storing data blocks which are likely to be accessed in the near future and thus allowing the disk to spin down for longer periods. However, if only a subset of the data on the disk is available on the flash memory then the disk may have to be woken up quite frequently due to cache misses or flushing. ...
... There has been quite a lot of research on saving energy in mobile storage systems by combining hard disks with flash memory in various configurations. March et al. [4], Bisson et al. [5], and Chen et al. [6] have all proposed using flash memory as a non-volatile cache, maintaining blocks which are likely to be accessed in the near future, and thus allowing a hard disk to spin down for longer time. Bisson et al. focused on the redirection of write requests to a flash memory device instead of a hard disk, while Chen et al. have recently studied a technique of partitioning the flash memory into a cache, a prefetch buffer, and a write buffer to save energy. ...
While hard disk drives are the most common secondary storage devices, their high power consumption and low shock- resistance limit them as an ideal mobile storage solution. On the other hand, flash memory devices overcome the main problems of hard disk drives, but they are still more expensive in the cost per bit over hard disk drives and can only support a limited number of erase cycles. In this paper, we show that combining the merits of a hard disk and a flash memory device can produce an energy-efficient secondary storage solution for mobile platforms. We propose an energy-efficient file placement technique for such heterogeneous storage systems. The proposed technique adapts an existing data concentration technique by separating read and write I/O requests. Experimental results show that the proposed technique reduces the energy consumption by up to 74.5% when the combination of a 1.8˝ disk and a flash memory is used instead of a single 2.5˝ disk, at the cost of small increase in the average response time.
Much attention has recently been put on the energy-saving scheme for heterogeneous drive (H-Drive) which combines SSD and HDD. This paper focuses on the energy-efficient file buffering schemes for H-Drive while ensuring disk's lifespan. We propose a frequency-energy based replacement scheme (FEBR for short) by adapting previous replacement algorithm FBR with the help of an energy-cost model. And based on the sliding-window scheme, we also present a self-adaptive disk power management scheme by taking the disk lifespan into account, which adjusts timeout threshold according to the statistical behavior of user accesses. To explore the applicability of the existing replacement schemes ranging from page-based to file-based buffering scheme, we evaluate their effectiveness on energy-efficiency, performance, and HDD lifetime and compare them with our proposed scheme. With extensive experiments on four real-world file usage traces collected in our office, some useful conclusions are drawn: energy-saving in H-Drive is feasible, it can reach as high as 70%~80%; FBR and its variant FEBR, and GDS are the best ones among all those online buffering schemes evaluated while FEBR has some advantages over FBR and GDS; the proposed self-adaptive disk power management scheme can effectively control the disk's lifetime and it is inappropriate to power disk on or off by using those fixed-timeout threshold scheme prevailed previously.
