3.6. Mobile data centers
One possible solution to the problem of retail network virtualization is the use of mobile data centers and rapidly deployable solutions. The mobile version of the data center is effective when the IT infrastructure is required to deploy at a remote site in the shortest possible time. A mobile data center can be used as a primary, backup or temporary. The market includes fast-deployed solutions (for example, Sun BlackBox, HP POD, IBM PMDC). The first mention of mobile data centers appeared in 2005.
Thus, APC implemented the mobile data center InfraStruXure Express Medium Density On-demand Mobile Data Center. The project was developed in the development of partners APC - Pacific Voice & Data (PVD Mobile Data Center) and IBM (Scalable Modular Data Center). According to IBM and APC, the Scalable Modular ata Center deployment takes 60-75% less time than building or upgrading a traditional server room, and its use saves significant design costs , the selection of components, the installation of access floors and operation of the data center.
3.7. The problem of large data and Virtualization
According to analysts, the amount of data on electronic media in the world is now within 1 & Ouml; 21-1022 bytes, and it is continuously increasing. Only for the period from 2010 to 2011 this volume increased 10 times. It is assumed that in 2012 the amount of data in the world will increase by 50% and will be about 2.7 • 102 'bytes. Despite the reduction in costs for the specific storage of data, there is a continuous increase in the cost of storing all the information. One of the reasons this analysts consider unreasonable data replication.
The Big Data problem (Big Data) is directly related to the problem of cloud computing, when data centers need to process huge amounts of information that are claimed by remote users. The data stored by enterprises can be varied in both types and terms of storage. Structured, partially structured and unstructured data in the form of textual information, structured database tables and information systems, documents, e-mail, music and multimedia files, structured data of trading systems with data on goods, suppliers, buyers and other contractors, etc.
When working with large data, the following basic problems arise.
1. The need to handle large amounts of data, approaching the maximum allowed for technical devices.
2. The speed of computer systems that support the processing of these data sets.
3. Variety of types of processed data and requirements for their relevance.
One of the solutions to such problems is the automatic separation of information, taking into account its relevance. Most of the data of the enterprise is of little or no demand once. At present, it becomes important to store, process and analyze unstructured information that has not been processed or deleted before. It can be stored on non-high-capacity data carriers, for example, on tape media. Often claimed data should be stored on devices with high I/O (read/write) speeds, for example SSD solid-state disks.
Related to the problem of "large data" High Performance Computing (HPC), conducted in analytical computing and distributed computing on multiple servers (grid computing). Currently, the concept of large data is associated with the changed requirements for their processing and analysis. The new requirements include:
• the need to process large data sets with a minimum delay, often in real time;
• quality decision making, which is made possible by processing heterogeneous information obtained from all available sources - structured and unstructured, internal and external, including from all kinds of sensors;
• the demand for devices capable of processing and transmitting data at a speed of several tens of gigabits per second;
• carrying out in-depth studies of data to identify dependencies, anomalies, compliance with the conditions of analysis, in order to obtain its qualitative model.
Due to processing difficulties, 80 to 90% of the data volume is unstructured. To handle large amounts of data, integrated solutions are used, combining data processing and analysis. New technologies for processing large data sets allow organizing their distributed storage and parallel processing in cluster structures. This is connected with the solution of such problems as localization and distribution of data between processors, load balancing, safe transfer of data from one computer to another in emergency operation modes, collection and aggregation of intermediate work results, etc.
The technologies for processing large data arrays include the Map Reduce model, the Apache Hadoop project, the Global File System (GFS) file system, the large-scale proprietary high-performance database, the XoSQL database, and the BI technologies.
The MapReduce model for batch processing, developed by Google, is relatively simple and easy to use. Corp. Teradata supplies the configured hardware and software solutions for the Aster MapReduce appliance to the data center, Oracle produces Oracle BigData appliance, and the EMC - solution Greenplum solutions appliance.
In-Memory Computing technology has recently been widely used when working with large data sets. However, it requires the use of powerful servers with multi-core processors and very large memory. This technology is more suitable for performing transactions with structured data, for example, article numbers, customer information, sales reports.
In addition, a technology is used for columnar storage and indexing of tables and processing of the entire database in RAM. It provides an increase in productivity by 10-1000 times.
In 2012, the world's major database management system vendors (DBMS) released updated software products designed to work with large data sets.
Sybase's company , which is part of the company SAP, has developed two databases: Sybase IQ 15.4 - for working with "large" data and Sybase Adaptive Server Enterprise (Sybase ASE) - for working with large data sets, which are processed by applications of SAP, in particular SAP Business Suite.
Sybase ASE for working with hot data is used by the real-time analysis tool - SAP HANA (SAP High-Performance Analytic Appliance). This is effective, for example, in the case of scalable blade servers with multi-core processors and adaptive algorithms for calculations. With scalability, SAP HANA can be applied to companies of all sizes.
This technology involves processing large amounts of data directly in the memory of the computer instead of storing it in static database tables. Based on the results of testing in 2011, the SAP HANA solution handles 10 thousand requests per hour. During this time, up to 1.3 TB of data is processed, and the results are displayed after a few seconds. For small and medium-sized businesses, the company offers a HANA Edge solution and an analytical module for resource planning of the Business One ERP.
Sybase IQ 15.4 > Sybase - is a single platform for analytically processing huge disparate arrays of structured, partially structured or unstructured data using various algorithms and analytical systems. Its main advantages are the columnbased storage of tables, a scalable service-oriented architecture, and high performance with low resource consumption.
We use the technology of mass-parallel processing, PlexQ, with the separation of all resources, which provides real-time service for thousands of users and analyzes large data sets. It allows solving complex analytical problems with acceleration from 10 to 100 times compared with traditional data stores. Clustering with the separation of all resources in combination with the column storage and the logical server mechanism provides high performance for all kinds of requests and load types for any data sets.
Microsoft SQL Server 2012 DBMS has acquired new tools and capabilities. So, the AlwaysOn function provides a high level of availability and reliability of the DBMS, and Xvelocity technology - high performance in performing analytical calculations. The advantage of Microsoft SQL Server 2012 is that all analytical tools are integrated with MS Office applications, MS Sharepoint, Dynamics class ERP systems, developed by Microsoft.
Simultaneously with the release of SQL Server 2012, Microsoft Corporation announced the use of xVelocity - the technology of computing in the memory of the computer (in-memory technologies). It provides high performance for data warehouses and business intelligence.
In a previously released SQL Server 2008 R2 application, Microsoft implemented a high-performance PowerPivot for Excel solution, also called Vertipaq. This technology is based on calculations in memory and column-oriented snoring in pei, as well as on the innovative method of data compression. In SQL Server 2012, Microsoft also used Vertipaq technology, which is an integral part of xVelosity.
Also We Can Offer!
- Argumentative essay
- Best college essays
- Buy custom essays online
- Buy essay online
- Cheap essay
- Cheap essay writing service
- Cheap writing service
- College essay
- College essay introduction
- College essay writing service
- Compare and contrast essay
- Custom essay
- Custom essay writing service
- Custom essays writing services
- Death penalty essay
- Do my essay
- Essay about love
- Essay about yourself
- Essay help
- Essay writing help
- Essay writing service reviews
- Essays online
- Fast food essay
- George orwell essays
- Human rights essay
- Narrative essay
- Pay to write essay
- Personal essay for college
- Personal narrative essay
- Persuasive writing
- Write my essay
- Write my essay for me cheap
- Writing a scholarship essay