PARALLAX - A NEW OPERATING SYSTEM

GAYA NAIR P I Sem, M.Tech, CSE Department, NCERC, Thrissur, Kerala ABSTRACT
Parallax, a new operating system, implements scalable, distributed, and parallel computing to take advantage of the new generation of sixty four bit multi-core processors. Parallax uses the Distributed Intelligent Managed Element (DIME) network architecture, which allows parallelism in resource configuration, monitoring, analysis and reconfiguration of the distributed software components. Parallax is implemented using the assembly language and provides a C/C++ programming application programming interface for higher level programming.

Introduction Sarathy et al.(2010) states that while analyzing the current application architectures the importance are provide to the continual measurement of productivity while managing resources. The two key emerging trends that point to the importance are, the execution and application management are becoming linked so that each application can be matched in real-time with distributed resources to meet its requirements. These real-time applications are to be adjusting with the resources available, changing workloads, business priorities, and latency constraints. The current network resource management control each resources for each application, is become more difficult. In the new class of multi-core architectures, hardwareassisted virtualization, and software strategies are used to exploit parallelism. The DIME computing model is different from these other approaches in that it introduces management and signaling capabilities to each computing element. Exploiting this computing model, a new operating system called Parallax is designed, which enables the execution of distributed and managed workflows within a server or across multiple servers. The new operating system also exploits fully the parallelism and multithread execution capabilities offered in these computing elements.

Distributed Intelligent Managed Element Network Goyal et al.(2009) states that the current cloud environments do not have great manageability and operability capabilities. The situation becomes more complicated. Operability is the ability to keep a service in a functioning and operating condition .This means that an element should be available, capable, reliable, safe and secure. Operability also requires the ability to determine the impact of change upon the services. Manageability is the ability to monitor and control the service creation, performance, scalability, availability, changes, risk and termination. Management of these two factors enabling flexible usage of the physical resources, dynamic orchestration and integration of management functions and also enabling corrective actions. Fault Configuration Accounting Performance and Security (FCAPS) have been adjudged to be the basic minimum management and operability capabilities necessary to offer useable services. Mikkilineni et al.(2009) states that the current definitions of cloud computing are just beginning to the basic foundation for cloud computing. Cloud computing can be defined as a pool of abstracted, highly scalable, and managed computing infrastructure capable of hosting end customer applications. In this layered model, each layer is responsible for different management functions. The network element layer implements logical entities within a device. The element management layer implements device level FCAPS management functions. The network management layer implements path management, topology management and fault isolation. The service management layer implements mechanisms to assure service level agreements and ensure quality of service. The business management layer implements strategic enterprise management functions, such as budgeting and billing. The DIME network enables FCAPS framework. This network architecture provides a direction to harness the power of many core servers with the high degree of scaling

and efficiency. It also eliminates many of the shortcomings of current solutions for solving the scalability issues in these systems. This network architecture allows the introduction of multiple instances of the operating systems in a single enclosure with socket communication among them. DIME network computing model integrates computational workflows with management workflows for providing dynamic realtime FCAPS management of distributed services. This network architecture also provides overlay of business workflow and its management workflow, service component encapsulation as an executable and these executable modules provides for the service to the sub-network model. End to end service network or sub-network management executed using the signaling network. Parallax Implementation Parallax is developed for sixty-four bit multi-core Intel processors. The DIME is implemented with two parts, intra and inter-DIME. This implementation decouples hardware infrastructure from services management. The service management console is used as a command console to control the workflow of services deployed using Parallax with command language interface. This operating system also provides the routing and switching functions to execute the services at the right destination. In a real implementation of a business workflow, the sub-network management implements using a DIME sub-network. Parallax provides memory management, central processing unit resource management, storage or file management, network routing and switching functions, and signaling management. Memory management is the act of managing computer memory. Memory protection is a way to control memory access rights on a computer. The main purpose of memory protection is to prevent a process from accessing memory that has not been allocated to it. In Parallax memory is assigned to a DIME on an as-needed basis. Each DIME maps to different pages in the linear memory system and cannot access pages to which it is not assigned. Security is provided at the hardware level for this

memory protection. After the completion of the execution the allocated spaces are returns to the system. Limits can be set on how much memory is able to allocate for itself. In CPU resource management the conventional operating systems allow for multiple programs or threads to share available CPU resources. This kind of sharing leads to a more complicated operating system that needs to take care of time slice management as well as internal context switching. Parallax allows for one program per core, that is the granularity is taken to an addressable thread level when multiple threads are supported by the core. Time slice management is not needed as each program running gets its own dedicated CPU core or thread. Performance of the program also increases over that of running on a conventional operating system, as context switching is no longer a factor. With dedicated resources, each network can be viewed as its own separate computing entity. After the completion of the task and it becomes free, it is given back to the pool of available resources. The Parallax Storage or file management offers local storage per server as well as centralized file storage shared via the orchestrator between all servers. Booting the OS via the network is also a possibility for systems that do not need permanent storages. In Parallax network management system all network communication is done over raw Ethernet frames. Conventional OS use transmission control protocol or internet protocol as the main communication protocol. By using raw packets, a much simpler communication framework is create, as well as removes the overhead of higher-level protocols, thereby increasing the maximum throughput. The use of raw Ethernet packets allow to use with the advanced technology attachment over ethernet protocol. The functionality of signaling management is routing or switching between inter and intra-DIME via the signaling and data channels. With the signaling layer, program parameters can be adjusted during run-time. DIMEs have the ability to communicate each other for co-operation as well as with the orchestrator for the purpose of compiling results. Instruction types can be directly encoded into the

sixteen bit ether type field of each raw-ethernet frame. By making use of the ethertype field, then streamline the way in which packets are routed within a system. The Parallax operating system currently using in new Intel laptops and servers which having parallel thread and multi-core features. This operating system has two components, intra-DIME FCAPS management methods that allow local management of fault, configuration, accounting, configuration, performance, and security management methods. Inter-DIME signaling methods that allow communication, coordination, and orchestration using the signaling abstractions that define the DIME network computing model. This concept implementation includes, fault management that provides a heartbeat of the network at a specified time interval, over the signaling channel, for implementing fault monitoring and recovery processes. Configuration management configures and controls the computing element to load, execute, suspend, stop, and recover tasks along with memory, storage, and network resource management. Account management keeps track of who requested the use of the computing resources as well as the duration. Performance management monitors the computing resources at set intervals and makes it available over the signaling channel. The security task specifies how the computing resources are accessed as well as assures that FCAPS tasks are executed only through authorized messages. Configuration of the DIME A major feature of the DIME computing model is its ability to implement a dynamically managed workflow using the sub networks. This involves instantiating and loading the network with appropriate computing tasks and assuring an end-toend connection management to assure the availability, reliability, performance, security and accounting management. Figure 1.shows the anatomy of a DIME network. The DIME orchestrator is used to manage this computing network using the signaling channel and to control the execution of computational workflow implemented utilizing the sub-network.

Figure 1. Anatomy of a DIME Network The node operating system part of the Parallax implements local network management including configuration of the managed intelligent computing element. This computing element implements the computing workflow component assigned to it. The network operating system part manages inter DIME communication, signaling and input output switching and routing functions. Different FCAPS functions define local policies such as broadcasting heartbeat, alarms, local security policy, performance parameter broadcast and utilization account details. Figure 2. shows a managed DIME network implementing a workflow.

Figure 2. A managed DIME network implementing a workflow. Workflow Configuration The components with the functionalities are stored in a service creation environment server. The workflow creation console uses the components and meta models to create a deployment workflow. This workflow loads and executes by the network across multiple physical servers based on the priorities, latency constraints and workload requirements. Each DIME is programmed to broadcast a heartbeat as its fault management function. A service management application can monitor the heartbeat and if it stops can initiate a service on a different network. When the heartbeat stops the service mediator instantiates the corresponding service in a different DIME based on policy specified. Each one is programmed to broadcast central processing unit utilization, memory allocation, input output bandwidth utilization, throughput utilization and storage capacity utilization as its performance management function. A service management application can monitor the performance parameters and can initiate or shut off services based on workflow scenarios. Accounting management keeps a call record of requests for managed intelligent computing element services. Security management allows various security schemes based on workflow requirements. The meta-model based FCAPS

management allows dynamic workflow management across the entire network based on end-to-end workflow requirements. The new computing model and the Parallax operating system implements node and connection management to deploy workflows as executable directed acyclic graphs. Conclusion In this paper presents an approach to the DIME network computing model that defines a method to implement a set of distributed computing tasks .The set of tasks implemented by the workflows arrange or organize in a directed acyclic graph, and executed by managed sub-networks. Each of the elements endows with its own computing resources to execute these tasks and have a management infrastructure to monitor and control the resources. Parallax implements by encapsulating each core with this network architecture. By implementing this network and its management using an orchestrator can demonstrated both fault and performance management automation. This model currently implements in current generation servers using features available in current operating systems such as ubuntu using the thread library.

References 1.Goyal, P., R. Mikkilineni and M. Ganti, (2009). FCAPS in the Business Services Fabric Model Proceedings of IEEE WETICE 2009, First International Workshop on Collaboration & Cloud Computing, Groningen, Netherlands, pp. 48-51. 2.Mikkilineni, R., V. Sarathy,(2009). Cloud Computing and Lessons from the Past, Proceedings of IEEE WETICE , First International Workshop on Collaboration & Cloud Computing, Groningen, Netherlands, pp. 57-62. 3.Sarathy, V., P. Narayan and R. Mikkilineni,(2010). Next Generation Cloud Computing Architecture: Enabling Real-Time Dynamism for Shared Distributed Physical Infrastructure,, 19th IEEE International Workshops on Enabling Technologies: Infrastructures for Collaborative Enterprises, wetice, pp.48-53