Hard disks are used in the data center for storage purpose. The data centers has progressed with cloud based data centers where it follows information life cycle management policy. Under these policies, the flow of data is maintained from one to another data centers . There has been significant progress in hard disk and their speed. There are large collection of hard disks at data center and this causes the I/O performance to decrease. The bottleneck is one of the problem that caused the I/O performance of data center to be low. The server side bottleneck are caused due to lack of memory or undersized I/O interface and CPU processing power. The lack of I/O interconnect bandwidth is one of the cause of the storage and I/O performance bottlenecks .
There are many aspects to look over data center which ranges from I/O performance to Hard disks. The investigation of all these are presented in this research paper along with the information life cycle management at data center .
Keywords: RAID (Redundant array of independent disks),
One of the fast growing services in the market of data centers is cloud based storage which follows information life cycle management policies. Every large data center has its own information life cycle management policies. Where as in cloud based data centers, there information are stored at several places for better security of data and its fast access by the customers . We will discuss the modern approach of data center towards the hard disk implementation. We are going to focus on IOPS (input output performance of hard disks), disks higher capability in cloud based data centers and issues of bottleneck in I/O performance. The development and research are being done to increasing the performance of modern data center and its input output server and application performances.
In the paper, we are going to investigate on datacenter and its information life cycle management in it, evaluation of hard disks at data center and its performance, I/O issues and its solution.
Hard disks at modern storage:
Cloud based data center uses numbers of hard disks that are arranged and organized into arrays. Array of hard disks are found in the cloud based data centers. The data that comes into the storage units are arranged into the disks in divided form. The spare disks are used for increasing the speed and performance of input and output . The RAID schemes are used to implement such concept. The hard disk controller controls the I/O Caching. The index are stored with the address to the files that increases the I/O performance . The use of Hot-swappable hard disks gives the removing option to hard disks from the array when it’s required. These technology are used in modern storage along with replication mechanisms which are faster and has been going through research. This mechanisms involves snapshotting which is storing of virtual memory into hypervisor-readable data that can be used for future.
Information life cycle management at data center:
The information has its life cycle at cloud based data centers. These information go through certain cycle and stored into the hard disks. The creation or discover is the first stage where the image, file, document are initiated. The second one is classify where these data are classified according to type and assigned to location for storage. The third stage is Archive/Retire and goes to encrypt . The data before storing are encrypted for number of security measures. e hold stage is storing of data into hard disk and last stage is retrieval and access of data and files that are being uploaded into cloud.
The main concern for the data in cloud and storage center with big data is I/O. It is not easy to access and store the data in faster way. There are number of issues that has been discovered which slows down the I/O performance . It is also dependent on disks. The seek distance is smaller when the platters are smaller which results in higher RPM, providing good Input output performance services.
4.1 Issues in I/O:
The issues that are faced in media system are late access of image, file storage redundancy and bottlenecks on server and hard disks. The bottleneck has impact on server, application and input output performance. When the graph of performance is obtained VS Disk I/O performance curve, there is increase with recent years. There is a big gap between processor and I/O performance that can be seen in the graph below:
There has been continuous need of more and more storage units in data centers over years with increase in users and data . This has increased more till 2016 as the research says.
Number of application uses these data which are industrial based and it is necessary to get data at fast rate. The impact of Bottlenecks is on this as well as it slows down the speed to data access resulting great loose of customer over media industry. The deadlock is another problem that slows down the data access in terms of response time. The data center uses additional measures of virus scan and purge and maintenance of data.
4.2 DEVELOPMENT in storage media technology to increase I/O performance.
The development in storage media technology to increase I/O performance is to reduce the bottleneck issues and make the data access faster. The data center has number of cache memory at local servers for fast access to data. The controller is used to handle requests for file using queuing technology. This decreases higher response time to certain level. The concept of big data and using modern technology has made data access much fast compared to last decade. The cloud computing has provided optimum capability for media technology. The technology of hot data is being used for fast access. In 2016 for accessing data faster rate, the use of flash storage could be implemented on the basic of cost per used capacity basis said Thomas Isakovich. The cache memory is provide for fast access of data which is made up of flash memory. The other ways for increasing performance could be block sizes, storage media and parallel accesses of disk subsystems.
When the data is to be accessed, there is transfer of data that takes place in blocks. With the increase in disk size, response time and throughput is found to be increased on linear basics in graph. When it comes to storage media, there is significantly large amount of data which are accessed by number of users at a time. The deadlock is occurred due to these, hence parallel access of hard disks is enabled. The controller handles such requests using queuing technology. This decreases higher response time to certain level. Over loading of disk subsystem is encountered with large number of parallel requests. The switching mechanism is implemented to overcome such problems. The controller uses these cache memories which was discussed in upper section . The different types of hard disks have different transfer protocol and it matters in I/O performance which are given below in the table:
In conclusion, we investigated into the evolution of media technology (storage) that are in the data center which supports I/O requirements . The hard disk was investigated in the report for faster I/O performance in cloud based data centers. The array of disks are used for storing data. The information life cycle of these data at cloud based data center was discussed and investigated closely. The low I/O performance on storage media technology is due to bottleneck at server and application level which could be more over years due to gap between processing and I/O performances . The cloud centers uses number of location of storing its data for fast access along with cache memory of that location. The recent research suggests that use of flash memory could be a solution to overcome the problem of I/O performance.
project management assignment help, business management assignment help, business management assignment, management assignment help services, strategic management assignment help, management accounting assignment, management homework help, marketing management assignment help, human resource assignment help, human resource management assignment help, managerial accounting assignment help, management accounting assignment help, financial management assignment help, it management assignment help, project management homework help, hospitality management assignment help, hr management assignment help, operations management homework help, brand management assignment help, database management homework help