The Compact Muon Solenoid (CMS) detector is one of the 4 detectors at the CERN LHC to study the Standard Model (SM) and even searching for signatures of new physics produced in high-energy collisions in the LHC. As many of the signatures in both SM and new physics include muons, it is crucial to possess high performance of muon identification and reconstruction. The performance of CMS muon identification and reconstruction is studied based on the run 2 data collected with the CMS detector at $\sqrt{s} = 13 ~\mathrm{TeV}$. The identification and reconstruction efficiencies are measured using tag and probe method. As High-Level Trigger (HLT) has made improvements in its algorithms during the run 2, we also highlight its upgraded feature by comparing with the previous algorithm.