NAM ANIMATOR USING NETWORK PROTOCOL

Transcription

1 International Journal of of Computer Engineering Engineering and Technology (IJCET), ISSN (Print), and Technology (IJCET), ISSN (Print) ISSN (Online) Volume 2 Number 1, Jan - April (2011), pp IAEME, IJCET I A E M E NAM ANIMATOR USING NETWORK PROTOCOL S.L.MOHAMMED SAJEER B.Sc (cs), M.Sc (IT), M, L.I.S., (Ind) 7A/1 Lake Road, Sammanthurai, Sri Lanka -ID: sajeer241@gmail.com ABSTRACT Protocol design requires Understanding state distributed across many nodes, complex message exchanges, and with competing traæc. Traditional analysis tools (such as packet traces) too often hide protocol dynamics in a mass of extraneous detail. This paper presents nam, a network animator that provides packet-level animation and protocol-specific graphs to aid the design and debugging of new network protocols. Taking data from network simulators (such as ns) or live networks, nam was one of the first tools to provide general purpose, packet-level and network animation. Nam now integrates traditional time-event plots of protocol actions and scenario editing capabilities. We describe how nam visualizes protocol and network dynamics. Keywords: Network protocol visualization, packet level animation, internet protocol design network simulation ns, nam. INTRODUCTION Network protocol designers face many difficult tasks, including simultaneously monitoring state in a potentially large number of nodes, understanding and analyzing complex message exchanges, and characterizing dynamic interactions with competing traffic. Traditionally they ve used packet traces to accomplish these tasks, but traces have two major drawbacks: They present an incredible amount of detail, which challenges the designer s ability to comprehend the data, and they are static, which hides an important dimension of protocol behavior. Although network simulators such as the VINT project s ns1 can easily generate numerous detailed Traces, they provide limited help for analyzing and understanding the data. The network animator that we developed in our work at the VINT Project provides packet-level Animation, protocol graphs, traditional time-event plots of protocol actions, and scenario editing capabilities. Nam benefits from a close relationship with Ns, which can collect detailed protocol. Information from a simulation. With some preprocessing, NAM can visualize data taken directly from real network traces. 61

2 Figure 1 NAM NAM BASED TECHNOLOGY Figure 2 NAM Description Gamida Cell identified sirtuins (SIRTs), which are type-iii lysine deacetylases of histones and of numerous transcription factors, as new stem cell targets. Gamida Cell's pre-clinical studies demonstrate that controlling the function of SIRTs with small molecules modulates stem cell fate, extends their ability to replicate and increases the number of blood-building stem cells known as hematopoietic stem cells. Based on these findings, Gamida Cell has developed a novel propriety technology for expansion of functional hematopoietic stem cells in in vitro cultures. The lead molecule of this technology is Nicotinamide (NAM), a Vitamin B3, a potent modulator of SIRT1 activity as well as a potent inhibitor of NAD+ -dependent ADP ribose transferees enzymes. Pre-clinical studies show the tremendous potential of this technology to increase stem cell migration, homing and engraftment efficacy, to rescue lethally irradiated primary and secondary recipients and to enhance myeloid and lymphoid reconstitution by maximizing the full therapeutic potential of bone marrow transplantation, tissue regeneration and gene therapy 62

3 Figure 3 NAM Based Technology NETWORK PROTOCOL VISUALIZATION Researchers have explored network protocol visualization in many contexts, beginning with static protocol graphs and visualization of large-scale traffic, and more recently including simulation visualizations and editors. Network visualization tools allow designers to take in large amounts of information quickly, visually identify patterns in communication, and better understand causality and interaction. Graphs of packet exchanges are useful for understanding cause and effect in complex protocols like TCP. Work at MIT2 and the University of Arizona3 is typical: Graphs show time against TCP sequence numbers on a2d graph, sometimes with annotations to show special events. Similar time-event graphs have proven useful in understanding reliable multicast behavior in SRM.4 Several groups have looked at visualization of large, static network data sets. Important questions include choice of layouts based on real-world geography or network topology and how best to use animation, color, and 3D. More generally, many researchers have tackled the problem of visualization of complex data.5systems like this share the principle that multiple linked views are essential for visualizing complex data. Network-specific visualization tools address this problem, allowing the user to take in large amounts of information quickly, to visually identifying patterns in communication, and to better understand causality and interaction. This paper presents nam, a net-work animator that provides packet-level animation and protocolspecie graphs to aid the design and de-bugging of new network protocols (Figure 1). Namwas one of the tools to provide general purpose, packet-level, network animation. Recent work has inte-grated traditional time-event plots of protocol actions and added scenario editing capabilities. Nam benetsfrom a close relationship with ns, the VINT 63

4 project s network ns [2] which can collect detailed protocol infor-mation from a simulation. With some pre-processing, nam can also be used to visualize data taken directly from real network traces. Packet Stack of headers and an optional data space are in the composition of NS packet. A packet header format is initialized when a simulator object is created, where a stack of all registered headers such as the common header that is commonly used by any object as needed, IP header, TCP header, RTP header and trace header is defined. The offset of each header in the steak is recorded. What this means is that whether or not a desired header is used, a steak consist of all registered headers is created. When a packet is allocated by an agent, and a network object can access any header in the steak of a packet it compiles using the linked off set value. However if u want to implement an application that talks to another application cross the network. You might use this feature with the little alteration. Another for this to create a new header for the application and upgrading the underlying agent to writhe data received from the application to the new header. Tools for NS simulation a. NAM b. AWK script language Figure 4 Packet Network Animator- NAM Visual interpretation of network topology is provided by the NAM. It was developed as apart of VINT project. Features given below. >Visual interpretation of network topology >Execution directly from the Tcl script. >Play, stop, fast forward, rewind, pause, display speed controller and a packet monitoring facility like Controls are used in NAM. >Information like throughput and number of packets on each link is present. >It also provides a drag and drop interface for creation of topologies. AWK Script 64

5 It is a test processing programming language. It is a utility for performing simple text processing tasks and it is a programming language for performing complex text processing tasks. Packet animation Animation allows the user to quickly see the status of each part of the network the top link is severely congested and dropping packets; the middle link is slightly busier than the bottom link and quickly compare algorithms the middle variation has one extra magenta packet, while the top versions ends many back-toback packets. Nam lets users adjust the animation speed and play it forward or backward, making it easy to find and examine interesting occurrences. The first step in a new animation is displaying the network topology. Network protocol has three different topology layout mechanisms to accommodate different needs: Automatic The default is an automatic layout algorithm based on a spring-embedded model.9the algorithm assigns attractive forces on all links and repulsive forces between all nodes and tries to achieve balance through iteration. Automatic layout can produce reasonable layouts of many networks without explicit user guidance, but it may not produce satisfactory results for complicated networks. If necessary, users can manually adjust the resulting layout. Relative For smaller topologies, relative layout is possible. The user species the relative directions of links (left, up, down). Nam places nodes relative to each other using link directions; link length is set proportional toils bandwidth and delay. Relative layout works very well for small topologies and has the desirable property that packet movement rate is consistent with link delay and bandwidth. The network in uses relative layout. Disadvantages of relative layout are that theuser must specify the directions of each link, that notall networks have a planer representation that satis-es delay constraints, and relative layout of a topology containing very di erent delays can result in very short links. Relative layout works well for small topologies and has the advantage that packet movement rate is constant and consistent with link delay and bandwidth. Relative layout s disadvantages are that the user must specify the directions of each link, not all networks have a planar representation that satisfies delay constraints, and relative layout of a topology containing different delays can result in very short links. Wireless Wireless layout associates each node with a physical location in a constrained area. Each node s 3D coordinate gives its position in the area (though visualization currently uses only two dimensions) and its velocity vector. Wireless visualizations typically lack explicit links. 65

6 Figure 5 Wireless Packet animation is straightforward once the topology is laid out. Trace events indicate when packets enter and leave links and queues. Packets are shown as rectangles with arrows at the front; queues as arrays of squares (see the left window of Figure 1). Packets can be colored based on codes set in the simulator or preprocessing to identify source and destination pairs. When queues packets are literally dropped, shown as small rolling squares falling to the bottom of the display Network statistics The animation component of nam only displays a sub-set of the simulation details present in the trace output. Additional information, such as packet headers or pro-tocol state variables, is handled by other nam components. The statistics component provides three ways to display this additional information. First, clicking on any of the displayed objects (e.g. packets and protocol agents) will bring out a oneshot panel showing object-specific information. Second, continuous monitoring of all available object-specific information maybe achieved by associating a monitor with entities of interest. Monitors remain associated with an object until explicitly removed by the user or until its under-lying object is destroyed. Other visualization methods In addition to packet animation, we experimented with other ways to visualize information. Users can specify node color and shape to indicate state such as membership in a multicast group use protocol agents to represent the state of a protocol instance at an end node, and display agents as small labeled rectangles attached to nodes. Protocol-specific graphs These graphs have long been used to understand TCP behavior and more recently to understand timer interaction in scalable reliable multicasts. Currently, protocol graphs for TCPand SRM, but we plan to add a pluggable API to support other, more-generic protocols event graphs for SRM (top right) and TCP (bottom center and bottom right). When a graph first comes up, a nam filter scans the trace file to extract the relevant information for a specific flow or protocol. 66

7 The advantage of integrating these views with is that it synchronizes graphs and packet animations. Moving a time slider or clicking on an interesting event in any view will update the time in all views. To help users coordinate events, displays each trace event in a consistent way through color or shape across views. Each trace event is displayed in the consistent way (i.e., col-or, shape, etc.) across views to help the user coordinate events. Scenario creation and editing We use nam in two complementary ways to assist scenario creation: With the scenario input facility that we recently added to nam, users can apply a traditional drawing approach to add nodes, links, and protocol agents. Nam then saves this scenario as an ns simulation script (in Tcl), which the simulator will process. The ns scenario generator uses nam to visualize large scenario topologies, using tools such as Georgia Tech s ITM10 to construct these scenarios. am uses automatic layout to present the topology to the user for acceptance or regeneration. CONCLUSIONS Nam development is ongoing, with a number of incremental improvements under consideration or planned. we plan to improve scenario-editing capabilities and add support for entering mobile node tracks. Two major focuses of future work remain: 1 st we would like to make nam easier to extend, providing better internal APIs to allow users to add custom controls to the output and to control object rendering. For example, one application would allow users to interactively control node colors to indicate application groups or characteristics. 2 nd we are just beginning to understand how to visualize large-scale networks (more than 100 nodes), and we intend to devote more work to this area. Network protocol visualization is easy to dismiss because its contributions to protocol development are indirect. More than a tool for fancy demos, visualization through nam can substantially ease protocol debugging and help developers understand dynamic behavior. REFERENCES 1. Brakmo, L. S., O'Malley, S. W., and e-treason, L. L. TCP Vegas: New techniques for Congestion detection and avoidance. In Proceed-ings of the ACM 2. T.J. Shepard, TCP Packet Trace Analysis, Tech. Report 494, Laboratory of Computer Science, Massachusetts Inst. of Technology, Cambridge, Mass., Comer, D., Internetworking with TCP/IP, Prentice Hall, Jacobson, V., Braden, R., Borman, A., TCP Extensions for High Performance, RFC 1323, May L. Breslau et al., Advances in Network Simulation, Computer, May 2000, pp Dempsey, B., Liebeherr, J., and Weaver, A., On Retransmission-based Error Control for Continuous Media Traffic in Packet-switching Networks, Technical Report CS-94-09, University of Virginia, Charlottesville, Virginia, Feb