Search

Informacijsko komunikacijska omrežja že nekaj časa rastejo in postajajo vse pomembnejša. Uporabljamo jih vsi in posegajo v vse pore sodobne družbe. S pomembnostjo in razvojem informacijsko komunikacijskih omrežij raste tudi odgovornost, ki je na strani tistih, kateri upravljajo z omrežji. Omrežja je potrebno dograjevati, jih nadzirati in učinkovito upravljati, da lahko uporabnikom zagotovijo kvalitetno in varno uporabniško izkušnjo. Zadnja leta opažamo pospešen razvoj orodij, ki lajšajo delo omrežnim inženirjem in zmanjšujejo vpliv človeškega faktorja. V magistrski nalogi obravnavamo načine avtomatizacije upravljanja in nadziranja omrežnih naprav. V nalogi so opisani različni omrežni protokoli in orodja za avtomatizacijo kot so NETCONF, RESTCONF, gRPC, Ansible, Chef in ostali. Poudarek je na načinu avtomatizacije z uporabo lastnih skript, napisanih v programskem jeziku Python in uporabi knjižnic Netmiko in NAPALM. Avtomatizacija v kontekstu IKT-omrežij predstavlja nadomestek človeške interakcije z omrežnimi napravami. Z avtomatizacijo se med sistemi izmenjujejo podatki, ki jih je potrebno obdelati, jih shraniti in jih interpretirati, da nam lahko podajo potrebne informacije. Zato sem v nalogi obravnaval možnosti hranjenja podatkov v relacijskih in nerelacijskih podatkovnih bazah in kritično vrednotil primernost obeh načinov pri avtomatizaciji. Poleg tega sem podrobno predstavil eno izmed orodij za učinkovito razčlenjevanje podatkov, to je Python knjižnica TextFSM. V nalogi je predstavljen primer, kako lahko več členov in orodij, ki skrbijo za različne funkcionalnosti, ki nam lajšajo delo z omrežij, združimo v celoto in jih povežemo v uporabnikom prijazno platformo. V primeru platforme gSmart, ki je opisana v magistrski nalogi, so za avtomatizacijo uporabljene lastne Python skripte, podatki se hranijo v graf bazi, upravljanje platforme pa je dogodkovno vodeno. Information and communication networks have been growing for some time now and are becoming increasingly important. We all use them, and they permeate every aspect of modern society. With the significance and development of information and communication networks, the responsibility of those who manage the networks also grows. Networks need to be expanded, monitored, and efficiently managed in order to provide users with a quality and secure user experience. In recent years, there has been accelerated development of tools that facilitate the work of network engineers and reduce the impact of the human factor. In my master's thesis, I address methods of automating the management and monitoring of network devices. The thesis describes various network protocols and automation tools such as NETCONF, RESTCONF, gRPC, Ansible, Chef, and others. The emphasis is on automation using custom scripts written in the Python programming language and the use of libraries such as Netmiko and NAPALM. Automation in the context of ICT networks represents a replacement for human interaction with network devices. Through automation, data is exchanged between systems, which needs to be processed, stored, and interpreted in order to provide us with the necessary information. Therefore, I explored the options for storing data in relational and non-relational databases and critically evaluated the suitability of both approaches for automation. Additionally, I provided a detailed presentation of one of the tools for effective data parsing, namely the Python library TextFSM. The thesis presents an example of how multiple components and tools that handle various functionalities, making our work with networks easier, can be integrated into a whole and connected to a user-friendly platform. In the case of the gSmart platform described in the master's thesis, custom Python scripts are used for automation, data is stored in a graph database, and the management of the platform is event-driven.

Distributed control systems are built on top of automation platform software. The system nodes might have different versions of the automation platform installed because some of them have been updated or they have been deployed later than the others. Thus, keeping system operable requires compatibility between the different automation platform versions. Compatibility must be verified with tests. The goal of this thesis was to start development of automated compatibility tests for automation platform. Test methods were looked from literature. Because there is not public research concerning automation platform compatibility testing, the methods were found from literature concerning other similar software fields. Based on theory, test cases can be included in components, they can be generated from system models, or they can base on user’s specifications. Compatibility verification tests can utilize captured execution data, reuse existing test assets, or test developers can write new targeted tests. Compatibility test system architecture consists of multiple interconnected nodes which are controlled and observed. The architecture can also have tools to analyse communication between the nodes. These method options are a starting point for a test development process. This thesis applied the above methods and presented a compatibility test development process for automation platforms. It starts from identification of the system enablers and limiters. They are used to select the most suitable methods. Then, the process continues with design of test cases and test architecture. The obtained test results are used in iterative development of the tests. The implemented tests noticed all known incompatibilities and some new issues were also found. Integration to continuous development of the automation platform was rapid because the tests reused existing assets. This thesis proved that the general compatibility testing methods work well also with automation platforms. The test development process helps to find the most suitable methods for each case.

Il rapporto descrive l'architettura della piattaforma di automazione concepita nel progetto Factory-ECOMATION, partendo dalle specifiche e dagli obiettivi dettagliati nel D7.1. Il documento espone aspetti teorici concepiti nell'ambito del progetto, quali l'introduzione dell'idea di avatar virtuale del sistema di manifatturiero e dell'architettura di automazione gerarchica e auto-simile a supporto dell'evoluzione dell'impianto verso un sistema ciber-fisico; la analisi degli algoritmi di controllo ottimo atti a governare il sistema grazie a tecniche predittive. Il report riporta anche aspetti più implementativi, legati all'ambiente di sviluppo della piattaforma di automazione, costruita sulla base della soluzione proposta da NXTcontrol, e gli strumenti a supporto del sistema di "virtual commisioning", quali l'adozione di strumenti di simulazione a eventi discreti.