Abstract: Techniques for distributed data storage and processing are described. In one embodiment, for example, a method may be performed that comprises presenting, by processing circuitry of a storage server communicatively coupled with a computing cluster, a first virtual data node to a distributed data storage and processing platform, performing a reliability evaluation procedure to determine whether the first virtual data node constitutes an unreliable virtual data node, and in response to a determination that the first virtual data node constitutes an unreliable virtual data node, performing a virtual data node replacement procedure to replace the first virtual data node with a second virtual data node. The embodiments are not limited in this context.