The international Square Kilometre Array (SKA) project to build two radio interferometers is approaching the end of its design phase, and gearing up for the beginning of formal construction. A key part of this distributed Observatory is the overall software control system: the Telescope Manager (TM). The two telescopes, a Low frequency dipole array to be located in Western Australia (SKA-Low) and a Mid-frequency dish array to be located in South Africa (SKA-Mid) will be operated as a single Observatory, with its global headquarters (GHQ) based in the United Kingdom at Jodrell Bank. When complete it will be the most powerful radio observatory in the world. The TM software must combine the observatory operations based at the GHQ with the monitor and control operations of each telescope, covering the range of domains from proposal submission to the coordination and monitoring of the subsystems that make up each telescope. It must also monitor itself and provide a reliable operating platform. This paper will provide an update on the design status of TM, covering the make-up of the consortium delivering the design, a brief description of the key challenges and the top level architecture, and its software development plans for tackling the construction phase of the project. It will also briefly describe the consortium’s response to the SKA Project’s decision in the second half of 2016 to adopt the processes set out by the Software Engineering Institute (SEI) for system architecture design and documentation, including a re-evaluation of its deliverables, documentation and approach to internal reviews.
Model Driven Engineering (MDE) as a key driver to reduce development cost of M&C systems is beginning to find acceptance across scientific instruments such as Radio Telescopes and Nuclear Reactors. Such projects are adopting it to reduce time to integrate, test and simulate their individual controllers and increase reusability and traceability in the process. The creation and maintenance of models is still a significant challenge to realizing MDE benefits. Creating domain-specific modelling environments reduces the barriers, and we have been working along these lines, creating a domain-specific language and environment based on an M&C knowledge model. However, large projects involve several such domains, and there is still a need to interconnect the domain models, in order to ensure modelling completeness. This paper presents a knowledge-centric approach to doing that, by creating a generic system model that underlies the individual domain knowledge models. We present our vision for M&C Domain Map Maker, a set of processes and tools that enables explication of domain knowledge in terms of domain models with mutual consistency relationships to aid MDE.
The SKA radio telescope project is building two telescopes, SKA-Low in Australia and SKA-Mid in South Africa respectively. The Telescope Manager is responsible for the observations lifecycle and for monitoring and control of each instrument, and is being developed by an international consortium. The project is currently in the design phase, with the Preliminary Design Review having been successfully completed, along with re-baselining to match project scope to available budget. This report presents the status of the Telescope Manager work, key architectural challenges and our approach to addressing them.
The Square Kilometre Array Telescope Manager (SKA TM) will be responsible for assisting the SKA Operations and Observation Management, carrying out System diagnosis and collecting Monitoring and Control data from the SKA subsystems and components. To provide adequate compute resources, scalability, operation continuity and high availability, as well as strict Quality of Service, the TM cyber-infrastructure (embodied in the Local Infrastructure - LINFRA) consists of COTS hardware and infrastructural software (for example: server monitoring software, host operating system, virtualization software, device firmware), providing a specially tailored Infrastructure as a Service (IaaS) and Platform as a Service (PaaS) solution. The TM infrastructure provides services in the form of computational power, software defined networking, power, storage abstractions, and high level, state of the art IaaS and PaaS management interfaces. This cyber platform will be tailored to each of the two SKA Phase 1 telescopes (SKA_MID in South Africa and SKA_LOW in Australia) instances, each presenting different computational and storage infrastructures and conditioned by location. This cyber platform will provide a compute model enabling TM to manage the deployment and execution of its multiple components (observation scheduler, proposal submission tools, MandC components, Forensic tools and several Databases, etc). In this sense, the TM LINFRA is primarily focused towards the provision of isolated instances, mostly resorting to virtualization technologies, while defaulting to bare hardware if specifically required due to performance, security, availability, or other requirement.
Large sensor-based science infrastructures for radio astronomy like the SKA will be among the most intensive datadriven projects in the world, facing very high demanding computation, storage, management, and above all power demands. The geographically wide distribution of the SKA and its associated processing requirements in the form of tailored High Performance Computing (HPC) facilities, require a Greener approach towards the Information and Communications Technologies (ICT) adopted for the data processing to enable operational compliance to potentially strict power budgets. Addressing the reduction of electricity costs, improve system power monitoring and the generation and management of electricity at system level is paramount to avoid future inefficiencies and higher costs and enable fulfillments of Key Science Cases. Here we outline major characteristics and innovation approaches to address power efficiency and long-term power sustainability for radio astronomy projects, focusing on Green ICT for science and Smart power monitoring and control.