Software incompatibility

Abstract: Architecture validation is becoming more and more important as diverging cost/performance criteria and competition cause the number of models within a computer family to proliferate. Some popular architectures are now being manufactured by many different companies and the chances of a company inexperienced with the architecture making mistakes is very high. Not only will errors in an implementation cause software incompatibility, the costs of fixing them are usually prohibitively high once there are a large number of defective machines in the field. Excellent evidence demonstrating the inadequacies of present testing techniques is implementation errors discovered in the field for many major computer families. This study was initiated in the hope that an error-oriented approach to architecture testing may provide a better detection of implementation errors.

Abstract: In the recent years, resource virtualization technologies of different types have been widely used in the ICT industry, particularly in cloud computing domain. These virtualization technologies can squeeze out hardware potential and consequently can significantly save CAPEX and OPEX. Virtualization technologies are adopted in network emulation to emulate heterogeneous nodes of target networks. At present, network emulators typically utilize legacy x86-based virtual machines (VMs) and light-weighted containers to emulate network nodes of heterogeneous architectures including ARM, SPARC, PPC, etc., which introduces software incompatibility to the original system software, protocols and applications, and may consequently jeopardize emulation fidelity. This paper focuses on mitigating the emulation incompatibility problem caused by node heterogeneity. Firstly, this problem is thoroughly investigated and analyzed on a cloud-based network emulation platform. Then, a hybrid virtualization approach concurrently utilizing multiple virtualization technologies, namely KVM, QEMU, and Docker to emulate heterogeneously architected nodes is derived and then implemented in the cloud-based network emulation platform. Functional verification and performance evaluation experiments have been carried out and have led to the conclusion that the elaborated hybrid virtualization approach can effectively dispose of the emulation incompatibility problem with affordable performance degradations. In addition, a case study of applying the hybrid virtualization approach to emulate a space-ground integrated network (SGIN) is conducted to show its effectivity and efficiency.

Abstract: Governments of different countries are increasingly expected to work together to address regional and global problems, provide disaster relief, satisfy international agreements, enable trade and movement of people across borders, and, in general, collaborate in achieving mutually-agreed goals. Information technology plays a key role in these transnational collaborations, particularly when country-specific information and IT resources are needed for each and every involved country to play their role in the collaborative efforts. In general, heterogeneity across IT systems is inevitable as it results from differences in economical and technical capabilities across countries, as well as from conceptual differences due to differences in agency missions and their regulatory context (which may, for example, specify what kind of software must be used) and differences in human IT resources. Integration may require use of new and/or existing hardware and/or software at different locations, processing and accessing data located in distinct agencies, and communication among many IT entities. In this context, heterogeneity can lead to several forms of incompatibilities among infrastructures, namely: (1) Hardware incompatibility: when machines do not have the expected architecture or capabilities to run machine-dependent software; (2) Software incompatibility: when needed software components require additional software that is either unavailable or conflicts with software used for local government functions; (3) Communication incompatibility: when multiple software components miss (a) interfaces for communication and (b) commonly understood protocols for communication; (4) Data incompatibility: when data maintained by software components use different organization, structure, semantic and natural language; and (5) Security and accessibility incompatibility: when integrating components with distinct or absent mechanisms for network security and access control.

Abstract: The article discusses the use of hybrid HPC clusters for the execution of software designed to calculate the electronic structure and atomic scale materials modeling. Modern software systems, which are designed to solve the problems of materials science, use the capabilities of various hardware computing accelerators to increase productivity. The use of such computing technologies requires the adaptation of application program code to hybrid computing architectures, which include classic central processing units (CPUs) and specialized graphics accelerators (GPUs). The use of large computing hybrid systems requires the development of methods for ensuring the workloading of such computing systems that will allow efficient use of computing resources and avoid equipment downtime. First of all, these methods should allow parallel execution of user applications using computational accelerators. However, in practice, software environments designed to solve application problems cannot be deployed in the same computing environment due to software incompatibility. In order to overcome this limitation and ensure the parallel execution of diverse types of materials science tasks, the creation of individual task execution environments based on virtualization technologies and cloud technologies. The continuation of virtualization technologies and the provision of cloud services is the construction of digital platforms. The article proposes the use of a digital platform for hosting scientific materials science services that provide calculations using various application software systems. Digital platforms make it possible to provide a unified user interface to scientific materials science services. The platform provides opportunities for finding the necessary scientific services, transferring source data and results between users, the platform and hybrid high-performance clusters.