Nowadays the growing usage of the various virtualization technologies force the necessarity of software-implemented network equipment (switches) as well. These softwer switches have many functions, capabilities, and with these abilities, a variety of software-based networks can be emulated. One of the most common representative of the technology is the open source Open Virtual Switch (OVS). Many multinational companies have been supporting the OVS development and use it in their products. One of the largest open-source cloud-based program is the OpenStack, where the OVS performs the full control of the network. Many special optimized version can be found, one of the best known is the Intel product, the OVDK (Open Virtual Development Kit).
By the evolution of virtualization technology there are growing needs to examine this switch interfaces' state to act in real-time response of network failures. Several methods exist to detect the problems. In my dissertation, I am presenting the OVS fault detection algorithms, such as Netlink, CFM (Connectivity Fault Management), and BFD (Bidirectional Forwarding Detection). In addition, I prepare my own fault detection application in OVS, which actively monitors the OVS database. Finally, I compare these algorithms and examine all in OVS and OVDK enviroment how efficiently they can operate.