Virtual Memory Paging and Segmentation - Article Example

? Full Paper Virtual Memory Paging and Segmentation Virtual memory virtual” means something that is not real) (Joseph, 2007) is an attribute of computer systems in which permanent data storages are utilized in various procedures, thus allowing RAM, or main memory, i.e. “active” or “fast” memory, to be free for other software (Virtual Memory, 2007). If the additional RAM is inserted, the computer system is then able to handle large files and operate with the help of “headroom” memory. Unfortunately, this kind of use of virtual memory is slower. Apart from many advantages, virtual memory and paging have various disadvantages as well. One of the challenges is the extra resource consumption, i.e. overhead for saving page tables in the memory (Null & Lobur, 2010). In environments where numerous applications are operating, the page table occupies a significant part of the physical memory. However, this challenge can be addressed by paging the page tables; this is a complex and confusing task (Null & Lobur, 2010). These challenges are not significant as compared to benefits of virtual memory and paging offers. Advantages of virtual memory and paging are easy to understand. For instance, programs are not limited to the physical memory present on the system, as virtual memory allows executing individual programs possessing larger virtual address space when compared to available physical memory (Null & Lobur, 2010). This process helps the programmers to avoid worrying about physical address space boundaries. Moreover, virtual memory also facilitates multi-tasking by executing more than one program simultaneously, and, as a result, system throughput and CPU utilization enhance. The limited size of pages and frames eases the tasks for placement and allocation for the operating system. Moreover, the paging process facilitates the operating system to protect and share the page on specific permissions (Null & Lobur, 2010). Apart from paging, segmentation is also known as the most common method for deploying virtual memory. This method incorporates a division of logical variable length segments (Dhamdhere, 2006) instead of allocating the virtual address space into fixed size and equal pages along with the division of physical address space into equal frames (Null & Lobur, 2010). Moreover, the physical memory remains intact and is not divided into any sort of fixed or equal size partition. Whenever there is a requirement of copying a segment in the physical memory, free memory is analyzed by the operating system, and after finding a chunk it is adequate enough to store an entire segment (Null & Lobur, 2010). Likewise, every segment possesses a base address that provides an indication of its location in the memory along with bound limits reflecting its size. Now every program incorporating various segments is now linked to a segment table instead of a page table. Likewise, the segment table incorporates the assembly of base for every segment (Jipping, 2007). The role of virtual memory in smart phone operating systems is also vital. For programming and designing virtual memory for a smart phone operating system, logical-segment addresses and physical-segment- addresses is the prime focus. Likewise, the segments are tagged with numbers, and these tags are incorporated within a logical address. For instance, a typical address format will be a , in which the segment number is the starting part of the address and the offset is identified as the last part of the address. Memory Management Technique for Paging on Distributed Shared Memory Framework The two analogous computer classes, i.e. shared memory and distributed memory, are symbolized by the DSM Systems (Hine, Mitrani, Tsur, & Gaines, 1979). This system works by providing shared memory generalizing in the system that further physically distributes the memory and thus combines the rewards of both the approaches. In fact, the DSM system is described as a memory system that physically distributes but is implemented rationally on a single shred addressed space (Hine, Mitrani, Tsur, & Gaines, 1979). In the middle of 1980’s, modern software of DSM implementation appeared (Hine, Mitrani, Tsur, & Gaines, 1979). This software provides the shared memory paradigm on top of message passing to the programmer. In general, this could be achieved in user level, run-time library, the operating system and programming language. (Hine, Mitrani, Tsur, & Gaines, 1979) On the contrary, for the typical DSM software implementations only support large grain size on the order of a kilobyte. This is due to the DSM management support that is accomplished only through virtual memory (Hine, Mitrani, Tsur, & Gaines, 1979). Therefore, if the data that is requested is not present in the local memory, a page fault handler indicates the fault and retrieves the page from the local memory of other node or disk of requesting node or another node. The software support for DSM is usually more elastic as compared to hardware support, thus enabling improved tailoring consistency method for the application performance. The first software DSM solution is proposed by the IVY. Here are the three categories into which the DSM software can be divided (Zelkowitz, 2000): Page-based. Variable-based. Object-based. Initially, these approaches are to answer the questions, i.e. where and how the transparency of remote access is? That will help to provide a better solution. Page-based The page-based software utilizes the memory management unit (MMU) that helps lock in remote access attempts. Variable-based This will need custom compliers in order to include particular information to the program code to detect any remote access. Object-based For this kind, a specialized programming language is required to verify a remote machine’s memory. The overall address space (spread over all the nodes) is divided into pages; this helps the virtual memory manager (VMM) to easily find an address space (Gopal, Beg, & Kumar, 2010). In fact, DSM manager is responsible to allow the remote machine to bring that page if the VMM needs it. This shows the page fault handling and a local page fault is revealed. In order to enhance the performance, reproducing of page can be done so that the transfer of the same page is not needed. The overall address space (consisting of physical memories of all the machines in the network) is represented by the DSM globally. This includes the programs that are needed for the machine execution. All the remote access information is detained by the DSM manager machine. Following are the choices of designs needed for DSM making (Gopal, Beg & Kumar, 2010): What kind of consistency model should the system provide? What kind of coherency protocol should be used? Where, with respect to the Virtual Memory Manager, does the DSM operate? What should be the granularity of the shared data? What should be the size of page? A Simple Linear Model of Demand Paging Performance A model of the patterns is needed for the predicting performance of an automatically projected multilevel memory system (Saltzer & Weissman, 1974). This will refer to the programs that provide information stored in the memory. In fact, some most recent investigational dimensions related to the Multics virtual memory propose that approximately a large number of simple program reference model be available (Saltzer & Weissman, 1974). This model states that the combined effects of the data and reference patterns with the effects of automatic management algorithm are needed to produce. The composite statement says the following: “The mean number of memory references between paging exceptions increases linearly with the size of the paging memory. The resulting model is easy to manipulate, and is applicable to such diverse problems as choosing an optimum size for a paging memory, arranging for reproducible memory usage charges, and estimating the amount of core memory sharing” (National Technical Information Service, United States, 2009). The Control of Response Times in Multi-Class Systems by Memory Allocation The inspection related to the opportunity of offering various services to jobs of different kinds via regulating memory is accomplished by the context present in computer systems (Hine, Mitrani, Tsur, & Gaines, 1979). In this situation the two parameterized algorithms that arrange main memory among two classes of jobs are measured (Hine, Mitrani, Tsur, & Gaines, 1979). At first, a study is conducted to determine optimal degrees of multi-programming and proportion of processor time by observing a closed system that contains a processor, paging and file devices along with fixed numbers of jobs. If the closed system is treated as a single server and a decomposed approach is implemented, the response times in an open system along with external arrivals are observed (Hine, Mitrani, Tsur, & Gaines, 1979). The purpose for this investigation is to know the effects of the memory allocation parameters that are expected for the response times under the two algorithms. As a result, numerical solutions and economical lower bounds that are expected in response time are observed (Hine, Mitrani, Tsur, & Gaines, 1979). Similarly, functions of the control parameters are obtained. The two jobs and more job classes are observed during the study. References Dhamdhere, D. M. (2006). Operating systems: A concept-based approach. McGraw Hill Higher Education. Gopal, B., Beg, R., & Kumar, P. (2010). Memory management technique for paging on distributed shared memory framework. International Journal of Computer Science & Information Technology, 2(2), 141-153. Hine, J. H., Mitrani, I., Tsur, S., & Gaines, R. S. (1979). The control of response times in multi-class systems by memory allocation. Communications of the ACM, 22(7), 415-424. Joseph, M. (2007) System software. Laxmi Publications. Jipping, M. 2007. Smartphone operating system concepts with Symbian OS: A tutorial guide. Chichester, England:; John Wiley & Sons. Null, L., & Lobur, J. (2010). Essentials of computer organization and architecture. Jones & Bartlett Learning. National Technical Information Service. (2009). Government reports announcements & index (Vol. 93). US Department of Commerce. Saltzer, J. H., & Weissman, C. (1974). A simple linear model of demand paging performance. Communications of the ACM, 17(4), 181-186. Virtual memory (2007). In Network Dictionary, 515-515. Zelkowitz, M. V. (2000). Emphasizing distributes systems . San Diego (Calif.): Academic press. Annotated Bibliography Dhamdhere, 2006. D. M. Operating systems: A concept-based approach. McGraw Hill Higher Education. Operating Systems, A Concept based Approach The book named “‘Operating Systems’ A Concept based Approach” provides comprehensive insights on memory segmentation and demonstrates the fundamental concepts behind this topic. The primary objective of the author of this book is to develop an understanding from the core fundamental concepts of the operating system and its semantics. As students are exposed to diverse concepts and techniques without practical knowledge of these concepts, the author tries to establish a general understanding. As the topic incorporates complex concepts, this source contributed to define memory segmentation in a simple and easy way for the reader. During my discussion on memory segmentation, I found this source extremely relevant for the reader to explain the size and length of memory semantics. Gopal, B., Beg, R., & Kumar, P. (2010). Memory management technique for paging on distributed shared memory framework. International Journal of Computer Science & Information Technology, 2(2), 141-153. Link: Database: Computers & Applied Sciences Complete Distributed Shared Memory Framework As distributed computing is now considered a fundamental concept for large enterprise computing environments, this article throws light on memory management in these distributed computing environments. In the literature section for memory paging and segmentation, the authors focus on distributed shared memory by introducing a framework. As distributed shared memory (DSM) has now become a popular paradigm in distributed computing, this article added value to the literature section. The introduction of DSM was highlighted along with the discussion of working semantics in a complex distributed computing environment. Hine, J. H., Mitrani, I., Tsur, S., & Gaines, R. S. (1979). The control of response times in multi-class systems by memory allocation. Communications of the ACM, 22(7), 415-424. Link: Database: Computers & Applied Sciences Complete Memory Management Technique for Paging on Distributed Shared Memory Framework The authors of this article describe the memory allocation in a distributed computing environment. This article contributed to the initial stages of memory management in a distributed computing environment. Along with the discussion of distributed shared memory evolution, analogous computer classes and shared memory generalizing are also discussed. As the article is quite old, the reader may first get the insights of the initial development of memory management and afterwards the recent technological discussion will come into place, making the flow of literature interesting. Joseph, M. 2007. System software. Laxmi Publications. Link: Book Virtual Memory Definition This source was specifically used for enhancing the definition of virtual memory in an easiest way for the reader. Likewise, a common definition resembling easy words is discussed in the beginning of the topic. To understand the topic is essential, as the reader will only develop interest if he or she understands the topic. Jipping, M. J. (2007). Smartphone operating system concepts with Symbian OS: A tutorial guide. Chichester, England: John Wiley & Sons. Link: Book Smart Phone Operating System Concepts As the cell phone industry has shown no limited potential, demand for high tech applications on smart phones is ever since increasing. This article discusses memory segmentation on operating systems running on smart phone architectures. As these phones have limited hardware to work on, logical-segment addresses and physical-segment addresses are discussed that are beneficial for mobile computing memory management. Null, L., & Lobur, J. (2010). Essentials of computer organization and architecture. Jones & Bartlett Learning. Link: Book Advantages and Disadvantages of Memory Paging and Segmentation This source provides a basis of the discussion on this topic. The book named Essentials of Computer Organization and Architecture comprehensively discusses advantages and disadvantages of memory paging and memory advantages of operating systems. Moreover, this source also takes a deep dive in the internal semantics of memory management. Along with discussing advantages and disadvantages, frames, pages and system performance enhancement techniques are also discussed. National Technical Information Service. (2009). Government reports announcements & index (Vol. 93). US Department of Commerce. Link: http://books.google.com.pk/books/about/Government_reports_announcements.html?id=Oa8bKUz4BHQC&redir_esc=y Memory Paging This source is a sub-part of the topic ‘a simple linear model of demand paging performance’. The discussion in this quoted text is associated with memory paging, as it focuses on references that are located in between paging exceptions and augment themselves along with the increase in the size of paging memory. We will discuss it briefly in the main source discussion. Saltzer, J. H., & Weissman, C. (1974). A simple linear model of demand paging performance. Communications of the ACM, 17(4), 181-186. Link: Database: Computers & Applied Sciences Complete A simple linear model of demand paging performance The authors of this linear model, Saltzer and Weissman, define and discuss the automatically projected multilevel memory system. As part of the literature, this source contributed a simple model that combines effects of the data along with reference patterns with the effects of automatic management algorithm that is considered as an output. The reason for selecting this source is that it demonstrates studies that have been performed for eliminated challenges and, consequently, making the virtual memory easily deployable and accessible. Virtual memory (2007). In Network Dictionary, 515-515. Link: Network Encyclopedia Definition of Virtual Memory Extracted from network encyclopedia, the definition is complete, precise, to the point and easy to understand for the reader. This is the basic aim of selecting this source and discussing it in the beginning of the topic. Zelkowitz, M. V. (2000). Emphasizing distributes systems . San Diego (Calif.): Academic press. Link: Book Three Categories for Distributed Memory Software This source further explains the Distributed Memory (DSM) concept and categorizes it into page-based, variable-based and object-based DSM. Relevant to the discussion, this source demonstrates new ways of virtual memory paging and segmentation by page-based virtual memory, which uses the memory management unit, variable-based virtual memory, which will need custom compliers in order to include particular information to the program code to detect any remote access, and object-based virtual memory, which requires a specialized programming language to verify a remote machine’s memory. Read More