High availability orchestration of Linux containers in mission-critical on-premise systems

Developers of mission-critical on-premise systems in the defence domain are looking towards containerization to tackle a number of challenges they are experiencing. Containers are often managed in clusters through orchestration technology, which is relatively novel. One important requirement for systems in the defence sector is the need for High Availability. Achieving High Availability using the novel orchestration technologies in mission-critical on-premise systems was not yet done. In this work, five architectures which fit mission-critical on-premise systems and theoretically pass the availability requirements have been designed. Through expert opinion a selection of two preferred designs was created. Prototypes of these designs were made and tested. These tests indicate that the designs are able to satisfy the availability requirements for mission-critical on-premise systems in the defence sector.