159 research outputs found
Universal Continuous Media I/o: Design And Implementation
No full textThe problem this paper addresses is how to modify an existing operating system's I/O subsystem to support new high-speed networks and high-bandwidth multimedia applications that will play an important role in future computing environments. The proposed I/O subsystem is called universal continuous media I/O (UCM I/O). This paper will cover the preliminary design of UCM I/O, some of the trade-offs and issues that need to be addressed in order to implement UCM I/O, and a a summary of work in progress. This work supported by ARPA, NSF, Ascom Timeplex, BNR, Goldstar, NEC America, NTT, Southwestern Bell, SynOptics, and Tektronix. UNIVERSAL CONTINUOUS MEDIA I/O: DESIGN AND IMPLEMENTATION Charles D. Cranor [email protected] +1 314 935 4203 Gurudatta M. Parulkar [email protected] +1 314 935 4621 1. Introduction As computers become more common and powerful, new applications are being proposed. Applications in areas such as distributed computing, imaging, and multimedia are becoming..
OpenVirteX: Make your virtual SDNs programmable
No full textWe present OpenVirteX, a network virtualization platform that enables operators to create and manage virtual Software Defined Networks (vSDNs). Tenants are free to specify the topology and addressing scheme of their vSDN, and run their own Network Operating System (NOS) to control it. Since OpenVirteX logically decouples vSDNs from the infrastructure, it also enables the introduction of features such as link and switch resiliency, and network snapshotting and migration of these tenant networks. OpenVirteX builds on the design of FlowVisor, and functions as an OpenFlow controller proxy between an operator's network and the tenants' network OSes. Our evaluations of this implementation show that i) OpenVirteX is capable of presenting tenants with configurable vSDNs while incurring a modest overhead to the control channel, and ii) that our architecture enables the introduction of features and enhancements such as link resilience to tenant networks
Real-time Upcalls: A Mechanism to Provide Real-time Processing Guarantees
No full textReal-time upcalls (RTUs) are an operating systems mechanism that can be used by applications to efficiently schedule code segments (or handlers) that must execute periodically. While the mechanism was conceived to support protocol processing with quality-of-service guarantees for networked multimedia applications it is general enough to be applicable in other domains like real-time image processing. Until now real-time threads have been the only mechanism for implementing protocols in user space with QoS guarantees. The RTU mechanism avoids the implementation complexity of the thread based approach while retaining its ability to ensure real-time behavior. In addition, our design simplifies protocol code, improves performance, and can be ported to most systems. A key feature of RTU scheduling is the pre-emption scheme that exploits the iterative nature of protocol processing by allowing an RTU to yield the CPU by returning from the invocation. This obviates the need for RTU handlers to..
Catching Up With the Networks: Host I/O at Gigabit Rates
No full textThe last few years have seen network data rates skyrocket from a few Mbps to a Gbps or more. However, a lack of integration of the host-network interface, the operating system, and network protocols has resulted in end-applications seeing only a small fraction of this total bandwidth being available for data transfer. The emergence of demanding applications in the realms of multimedia and virtual reality provides further impetus in the drive to overcome this problem. In this paper, we present the design of a high performance ATM host-network interface for workstations and servers that can support a bidirectional sustained data rate in excess of a gigabit per second. A prototype of the interface is being built at Washington University as part of an ARPA-sponsored gigabit local ATM testbed. Our interface design, which emphasizes seamless integration with the OS and network protocols, features: support for streaming data from I/O devices (e.g., cameras, disk arrays, etc.) to the network or vice-versa, as well as from device-to-device, while bypassing the main system bus; an ATM interconnect that extends to the desk-area; a zero-copy interface to system memory that is achieved through the use of page remapping techniques; full AAL-5 segmentation and reas-sembly; pacing control that provides for single-parameter bandwidth reservation; a high degree of scalabil-ity in terms of the number of I/O devices that can be simultaneously supported; low-cost (one ASIC); and multiprocessor support
Charismata and compassion : Dhinakaran, Charismatic healing and Pastoral Pentecostalism in South India : a practical theological assessment
No full textThis dissertation is a theological assessment of the Jesus Calls divine healing movement as it developed in South India. It analyses the thought of the founder D. G. S.
Dhinakaran, India's pre-eminent healing evangelist and explores its potential and adequacy as a pastoral theology. It draws out Dhinakaran's healing theology and model for ministry that has significantly contributed to Indian mission and attained international recognition. Until now, this movement has received no critical analysis.
The author, a native of South India, as an observer-participant, takes up this task.
The emergence and impact of the modem Pentecostal movement and the `full gospel' with healing as its flagship is discussed. The study highlights lesser-known precursors
and theological roots that give Indian Pentecostalism its distinct identity from the 1906 Azusa Street Revival, USA. It reveals how Charismatic Christianity flourishes due to
its bhakti spirituality, guru leadership, apostolic charismata and practical compassion. The thesis maintains that the dialectic of Pentecostal power and pastoral care is an effective mission strategy and proceeds to demonstrate this in Dhinakaran's ministry.
The research method progressed from a description of Dhinakaran's healing praxis via theological analysis to a critical assessment. It shows some major influences and the
appeal of his prayer movement and argues that the key to its success lies in its highly pragmatic, culturally adaptive and syncretic nature. Three doctrinal concepts that are embodied in Dhinakaran's model are presented: compassion, the wounded healer and healing evangelism, which correlate within a theological apparatus to make
interpretative sense of his praxis resulting in what is termed `Pastoral Pentecostalism'.
The hybridity in Dhinakaran's `miracle healing' is explored within two wider contextual interfaces: the traditional Hindu culture with the shamanic manthiravadi and an imported but adapted version of American faith healing. Here, an easternwestern synthesis is shown to contribute at once to a viable indigenous ministry and to global trends in Charismatic Christianity, allowing each to inform and shape the other. Dhinakaran's distinct prosperity message is located within the Guru movement and assessed against the American health-wealth gospel for pastoral integrity. The thesis advocates a Pastoral Pentecostalism that holds charismata and compassion in creative tension and re-presents Dhinakaran as a significant charismatic healing evangelist
Building Quality Assurance into Metadata Creation: an Analysis based on the Learning Objects and e-Prints Communities of Practice
No full textThis paper challenges some of the assumptions underlying the metadata creation process in the context of two communities of practice, based around learning object repositories and open e-Print archives. The importance of quality assurance for metadata creation is discussed and evidence from the literature, from the practical experiences of repositories and archives, and from related research and practices within other communities is presented. Issues for debate and further investigation are identified, formulated as a series of key research questions. Although there is much work to be done in the area of quality assurance for metadata creation, this paper represents an important first step towards a fuller understanding of the subject.
Efficient User space Protocol Implementations with QoS Guarantees using Real-time Upcalls
No full textReal-time upcalls (RTUs) are an operating systems mechanism to provide quality-of-service (QoS) guarantees to network applications, and to efficiently implement protocols in user space with (QoS) guarantees. Traditionally, threads (and real-time extensions to threads) have been used to structure concurrent activities in user space protocol implementations. However, preemptive scheduling required for real-time threads leads to excessive context switching, and introduces the need for expensive concurrency control mechanisms such as locking. The RTU mechanism exploits the iterative nature of protocol processing to eliminate the need for locking, and reduce asynchronous preemption, while ensuring real-time operation. In addition to efficiency, eliminating the need for concurrency control considerably simplifies protocol code. RTUs have been implemented in the NetBSD OS on the Sparc and Pentium platforms. We used the RTU mechanism to implement a sender and a receiver program that communicate using UDP sockets over 155Mbps ATM, and compared the throughput with that obtained when RTUs are not used. Our results show that the RTU based programs maintain the same throughput regardless of other system activities. Without the RTU mechanism, background load reduces throughput by as much as 80%. Using the RTU mechanism, total network bandwidth can be partitioned among different UDP streams, and delivered to user programs even with background system load. We have also implemented the TCP protocol in user space using the RTU mechanism. For each TCP connection, RTUs are setup for performing TCP output, input, and timer processing functions. The use of RTUs, in conjunction with shared memory between kernel and user processes for data movement, and the ability of the ATM adaptor driver to separate headers and data, makes our user level TCP implementations the most efficient one that we know of for providing QoS guarantees within the endsystem
Efficient Quality of Service Support in Multimedia Computer Operating Systems
No full textThis report describes our approach towards providing quality of service (QoS) guarantees for network communication within the endsystems to support multimedia applications. We first address the problem of QoS specification by identifying a set of application classes and their QoS parameters that cover the communication requirements of most applications. We then describe the QoS mapping problem, and show how requirements for resources (such as the CPU, the network interface adaptor and network connections) can be automatically derived from the application QoS parameters. We then deal with the QoS enforcement issue in which we describe techniques for scheduling protocol processing threads in order to reduce context switching overhead, as well as derive sufficiency conditions in order to provide predictable performance. We integrate all these solutions in a protocol implementation model. The key feature of the model is that protocols are part of the application process and are processed using protocol threads with individual scheduling attributes derived using our QoS mapping method. We propose several performance improvement techniques for application level protocol implementations that can reduce the high cost of data movement and context switching in these implementations. A significant component of this work will consist of implementation and experimentation which will result in significant contributions of practical utility
- …
