Scalability issues for mass multimedia storage systems

Customer requirements have guided the evolution and implementation of software and hardware computer data storage technologies. The concept of RAID originated with a group of researchers at the University of California at Berkeley (UCB). They were investigating the use of an array of small capacity, inexpensive personal computer direct access storage devices (DASD) managed by a special type of a controller, as an alternative to the single large expensive disk (SLED) subsystems more common on mainframe computers. The disk arrays can be configured differently to provide different levels of operational capabilities, each providing special advantages in reliability and data availability, speed and economy. Different configurations are referred to as different levels of RAID. Berkeley originally defined a basic set of RAID levels ranging from RAID, level 0 (RAID 0) through RAID, level 5 (RAID 5). The RAID technology vendors have since extended the RAID level domain by defining RAID 6 and beyond to accommodate additional functionality. Each RAID type has shown itself to be suitable for certain applications. The projected reduction in storage costs and facilities, $/GB and ft2/GB respectively, using the present RAID technology and its projected evolution are phenomenal.

Distributed i>redundant array of inexpensive disks (RAID) is often embedded in a cluster architecture. In a centralized RAID subsystem, the false sharing problem does not exist, because the disk array allows only mutually exclusive access by one user at a time. However, the problem does exist in a distributed RAID architecture, because multiple accesses may occur simultaneously in a distributed environment. This problem will seriously limit the effectiveness of collective I/O operations in network-based, cluster computing. Traditional accesses to disks in a RAID are done at block level. The block granularity is large, say 32 KB, often resulting in false sharing among fragments in the block. The false sharing problem becomes worse when the block size or the stripe unit becomes too large. To solve this problem, we propose an adaptive sector grouping approach to accessing a distributed RAID. Each sector has a fine grain of 512 B. Multiple sectors are grouped together to match with the data block size. The grouped sector has a variable size that can be adaptively adjusted by software. Benchmark experiments reveal the positive effects of this adaptive access scheme on the performance of a RAID. Our scheme can reduce the collective I/O access time without increasing the buffer size. Both theoretical analysis and experimental results demonstrate the performance gain in using grouped sectors for fast access of distributed RAID.

Redundant Arrays of Inexpensive Disks (RAID) is a popular technique used to improve the reliability and performance of secondary storage. Of various levels of RAID discussed, RAID1 and RAID5 have become more popular. Mirroring or RAID1 maintains multiple copies of the data, generally provides best performance and is easier to configure. Rotating parity scheme or RAID5 is the least expensive RAID scheme with good large update performance. It suffers from poor small update performance and performance drops sharply when a diskfails and the array enters degraded mode. Configuring RAID5 is more involved. This paper presents the design and implementation of a host-based driver for a multi-tier RAID storage system, currently with 2 tiers: a small RAID1 tier and a larger RAID5 tier. Based on access patterns, the driver automatically migrates frequently accessed data to RAID1 while demoting not so frequently accessed data to RAID5. The prototype provides reliable persistence semantics for data migration between the tiers using ordered updates. Mechanisms are separated from policies through an API so that any desired policy can be implemented in trusted user processes. Finally, we present comparison of the performance of our system with comparable systems using striping and RAID5.

HDDs are like the bread in a peanut butter and jelly sandwich—sort of an unexciting piece of hardware necessary to hold the “software.” They are simply a means to an end. HDD reliability, however, has always been a significant weak link, perhaps the weak link, in data storage. In the late 1980s people recognized that HDD reliability was inadequate for large data storage systems so redundancy was added at the system level with some brilliant software algorithms, and RAID (redundant array of inexpensive disks) became a reality. RAID moved the reliability requirements from the HDD itself to the system of data disks. Commercial implementations of RAID range from n+1 configurations (mirroring) to the more common RAID-4 and RAID-5, and recently to RAID-6, the n+2 configuration that increases storage system reliability using two redundant disks (dual parity). Additionally, reliability at the RAID group level has been favorably enhanced because HDD reliability has been improving as well.

This paper is a summary and evaluation of redundant array of independent disks (also called redundant array of inexpensive disks before) that was proposed years ago. Recently more information is produced every year, and the demand of data storage is increasing exponentially due to Moore's Law. So, it's much more common to group a dozen of hard disk drives together, to achieve better performance or to achieve better data safety. Firstly, the redundant array of independent disks aims to solve the data safety/data in-out performance problems, and multiple ways are proposed including Level 0 to solve the performance problem sacrificing system stability and data safety and Level 1 aims to achieve better stability but using only 50% (or less) of the data capacity. The development of redundant array of independent disks tends to be merged and have more functions in one single system of redundant array of inexpensive disks.

The authors discuss various types of RAIDs (redundant arrays of inexpensive disks), a cost-effective option to meet the challenge of exponential growth in the processor and memory speeds. They argue that the size reduction of personal-computer (PC) disks is the key to the success of disk arrays. While large arrays of mainframe processors are possible, it is certainly easier to construct an array from the same number of microprocessors (or PC drives). With advantages in cost-performance, reliability, power consumption, and floor space, the authors expect RAIDs to replace large drives in future I/O systems. >

RAID (Redundant Array of Inexpensive Disks) are widely used in storage servers. Level-5 RAID is one of the most popular RAID architectures. Numerical analysis of exact Markovian dependability models of level-5 RAID architectures with orthogonal organization is unfeasible for many realistic model parameters due to the size of the resulting state space. In this paper we develop approximate dependability models for a level-5 RAID architecture with orthogonal organization which have small state spaces. We consider two measures: the steady-state unavailability and the unreliability. The models encompass disk hot spares and imperfect disk reconstruction. Using bounding techniques we analyze the accuracy of the models and show that the models are extremely accurate.

Storage networks is a popular solution to constraint servers in storage field. As described by Gibson's metrics, the performance of multi-level networked RAID (redundant arrays of inexpensive disks) based on cluster is almost the same to that of improved 2D-parity. Compared with other schemes, it is lower cost and easier to realize.

Flash memory based solid state drives (SSD) are increasingly common in portable and mobile computing devices such as laptops, mobile phones, and tablets. Due to space, weight, and power constraints, portable devices are often restricted to a single storage device, which makes them susceptible to data loss from internal errors. On the other hand, mission-critical mobile applications like wireless healthcare always demand a high level of data reliability. This is mainly because data sampled from mobile and dynamic environments are most likely irreproducible. An effective approach to improving storage and data reliability is the RAID (redundant arrays of inexpensive disks) organization. However, the multiple disks required to implement RAID make it incompatible with the aforementioned restrictions of many portable devices. In this paper, we propose a SIRF (single internally redundant flash) architecture that leverages the internal hierarchical structure and parallelism of SSDs to provide redundancy similar to RAID in a single drive configuration. The initial effort focuses on implementing SIRF-1 (mirroring), which is the corollary to its RAID-1 counterpart. In SIRF-1, data is mirrored across SSD channels to optimally exploit parallelism for both read and write operations. Simulation results show that for read-dominant workloads SIRF-1 significantly outperforms a non-mirrored SSD by up to 39.5% in terms of mean response time. For write-intensive workloads, SIRF-1 pays a performance penalty no more than 5.5%.

This paper introduces the concept of Coding Over Graphs: given an arbitrary graph, assign information to vertices of the graph in such a way that the information can be retrieved at every vertex by exploring its close proximity. This framework is a generalization of RAID (Redundant Array of Inexpensive Disks). We study a specific coding over graphs scheme which has the property that as the number of failures increases, the information can be retrieved by exploring slightly larger proximities—therefore a graceful trade-off between reliability and performance is provided. An efficient algorithm for that scheme for the graphs being trees is presented. The algorithm can always find a ‘perfect’ coding over trees scheme.

Redundant array of inexpensive disks (RAID) based storage systems performance is limited to the sequential nature of the central processing unit (CPU), and they consume high amounts of energy because they need erasure coding for striping data and parity into storage devices. This paper proposed an energy-aware GPU based scheduling. The proposed scheduler differs from existing RAID in that it can reduce the number of CPU cycles and coding time. The proposed system generates parity by using a GPU, stripes parity at the initiator server and stripes data at the target server. We also apply an energy-aware scheduling scheme based on solid state disk-based data storage and hard disk drive-based parity storage. Experimental results show that the energy consumption by GPU-RAID is 45% less than Linux-RAID.

<div class="htmlview paragraph">Almost every automotive noise, vibration, and harshness (NVH) engineer who has ever looked at a fast Fourier transform knows the difference between instruments with 90 dB and instruments with 120 dB of dynamic range. NVH engineers understand instrumentation specs such as 24-bit resolution analog-to-digital converters and alias-free signal bandwidth. However, with modern noise and vibration instrumentation systems now being almost completely built on the PC, these specs neglect the most important X factor: the PC itself.</div> <div class="htmlview paragraph">No other aspect can affect the performance of an instrumentation system for a sensor array more than the components of the PC. Fortunately, a variety of off-the-shelf PC technologies built on industry standards are available to make it easier and less expensive than ever before to instrument and manage data from large systems. But an NVH engineer must wade through a sea of options to choose the right PC technologies for desired instrumentation system performance.</div> <div class="htmlview paragraph">This paper covers a variety of PC topics including chipset architecture, hard drive technologies such as redundant array of inexpensive disks (RAID), and bandwidth and latency limitations of internal and external buses such as Gigabit Ethernet, PCI Express, and ExpressCard.</div>

An algorithm for redundant management of distributed data using a minimal amount of data replication is described and analyzed. The main results on storage space utilization, I/O (input/output) performance, and reliability are outlined. The recently introduced RAID (redundant array of inexpensive disks) concept is extended to a distributed computing system. The resulting distributed storage architecture, called RADD (redundant array of distributed disks), is shown to support redundant copies of data across a computer network at the same space cost as RAIDs do for local data. Thus, a RADD increases data availability in the presence of both temporary and permanent failures (disasters) at local sites. During normal operation, the RADD scheme offers performance comparable to the two-copies systems. As such, RADDs should be considered a possible alternative to traditional multiple-copy techniques as well as to other high-availability schemes. >

Private Information Retrieval (PIR) allows to privately request a block of data from a database such that no information about the queried block is revealed to the database owner. With the rapid rise of cloud computing, data is often shared across multiple servers, making multi-server PIR a promising privacy-enhancing technology.In this paper, we introduce RAID-PIR, an efficient and simple multi-server PIR scheme, which has similar approach to RAID (Redundant Arrays of Inexpensive Disks) systems. Each server stores only a part of the database, its computational complexity depends only on this part, and multiple blocks can be queried efficiently in parallel. RAID-PIR improves efficiency over known PIR protocols, using only very efficient cryptographic primitives (pseudo-random generator). We demonstrate that RAID-PIR is practical and well-suited for cloud deployment as it reduces the communication as well as the computational workload per server.

Summary form only given. We present an implementation of a VOD (video on demand) system that addresses the problem of increasing video stream supply capabilities. To obtain fundamental video stream transfer characteristics on Ethernet, we first performed an experiment in which it became clear that bottlenecks of the CPU of a video server were the most serious. To solve this problem we implemented a distributed RAID4 (Redundant Arrays of Inexpensive Disks) method video server. Distributed RAID means extending the RAID method usually applied to disks to the servers, so as to achieve improvements in performance and reliability. The great advantage of this architecture is that the linear performance improvement can be achieved by adding inexpensive servers of a diffusion model.