nÃo hà / Devido ao crescente uso de computadores e a forte dependÃncia dos mesmos em vÃrias aplicaÃÃes, as UPS (Uninterruptible Power Supplies) vÃm adquirindo cada vez mais importÃncia. Estes dispositivos encontram-se muitas vezes dentro de um contexto de um sistema distribuÃdo, o que fortalece a necessidade de gerenciamento dos mesmos. Devido à grande utilizaÃÃo das UPS em computadores, e devido ao fato destes, por sua vez, estarem inseridos em um ambiente de redes TCP/IP, torna-se bastante viÃvel a idÃia de se usar um sistema de gerenciamento que seja compatÃvel com esse ambiente, e que faÃa uso de suas potencialidades. No presente trabalho à apresentado a implementaÃÃo de um sistema de gerenciamento SNMP (Simple Network Management Protocol) para UPS que possibilita o gerenciamento de um nÃmero arbitrÃrio de UPS, e que permite a autenticaÃÃo remota para gerentes, facilitando o uso para um maior nÃmero de dispositivos. Detalha-se a implementaÃÃo do sistema bem como os testes realizados. No final do trabalho à apresentada uma proposta para trabalhos futuros na Ãrea. / The reliability of a computer system is based not only on the computer itself, but also on its related power system. For that reason, more and more critical systems are counting on UPSs (Uninterruptible Power Supplies) to account for a continuous supply of power. Although the task of managing an UPS for a standalone system be quite straightforward, when it comes to computer distributed systems that task has to follow an equally distributed approach. Depending on the strategy used to manage that distributed UPS network, this task may be very complex. However, keeping in mind that most computer systems in a distributed architecture make use of TCP/IP to be logically connected, one could profit from the facilities provided by this protocol and design a management strategy that makes use of it. In fact, TCP/IP is not the only suite of protocols used to interconnect computer systems. SNMP (Single Network Management Protocol) is also frequently used as a way of managing (hardware and/or software) devices associated with computers. This dissertation proposes and implements a way of managing distributed UPSs by using SNMP, taking advantage of all the facilities furnished by the protocol. This works describes how the use of SNMP makes it possible to manage an arbitrarily number of UPSs remotely. This makes the design of UPS management systems a relatively easy task, and contributes to introduce a high degree of flexibility on the task of managing distributed UPSs. Finally, it is described how the system and tests were implemented as well as proposals for future work is presented.

A human brain has billions of processing elements operating simultaneously; the only practical way to model this computationally is with a massively-parallel computer. A computer on such a significant scale requires hundreds of thousands of interconnected processing elements, a complex environment which requires many levels of monitoring, management and control. Management begins from the moment power is applied and continues whilst the application software loads, executes, and the results are downloaded. This is the story of the research and development of a framework of scalable management tools that support SpiNNaker, a novel computing architecture designed to model spiking neural networks of biologically-significant sizes. This management framework provides solutions from the most fundamental set of power-on self-tests, through to complex, real-time monitoring of the health of the hardware and the software during simulation. The framework devised uses standard tools where appropriate, covering hardware up / down events and capacity information, through to bespoke software developed to provide real-time insight to neural network software operation across multiple levels of abstraction. With this layered management approach, users (or automated agents) have access to results dynamically and are able to make informed decisions on required actions in real-time.