In the Performant and Available Computing Systems (PACS) Lab, we analyze, model, and promote the non-functional properties, performance and availability, in particular, of computing software systems. The essential innovation is to make these systems smart enough to guarantee their intended operation with the expected quality of service. To this end, rigorous analytical modelings using queuing theory, stochastic processes, machine learning, optimization techniques, control theory, and Petri net may be leveraged. Large-scale validation of analytical models will be performed on cloud computing and edge environments.