Galactic bulges of early-type spirals and elliptical galaxies comprise primarily old stars, which account for more than half of the total stellar mass in the local Universe. These stars collectively generate a long-lasting feedback via stellar mass loss and Type Ia supernovae. According to the empirical stellar mass loss and supernova rates, the stellar ejecta can be heated to more than 107 K, forming a very hot, diffuse, and ironrich interstellar medium. Conventionally a strong galactic wind is expected, especially in low- and intermediate-mass early-type galaxies which have a relatively shallow potential well. X-ray observations, however, have revealed that both the temperature and iron abundance of the interstellar medium in such galaxies are unexpectedly low, leading to the so-called “missing feedback” and “missing metal” problems. As an effort to address the above outstanding issues, we have carried out a series of hydrodynamic simulations of galactic bulge feedback on various scales. On galactic halo scales, we demonstrate that the feedback from galactic bulges can play an essential role in the halo gas dynamics and the evolution of their host galaxies. We approximately divide the bulge stellar feedback into two phases: 1) a starbusrtinduced blastwave from the formation of the bulge built up through frequent major mergers at high redshifts and 2) a gradual feedback from long-lived low mass stars. The combination of the two can heat the surrounding gas beyond the virial radius and stop further gas accretion, which naturally produces a baryon deficit around Milky Way-like galaxies and explains the lack of large-scale X-ray halos. On galactic bulge scales, we study the collective 3-dimensional effects of supernovae with their blastwaves resolved. We find that the sporadic explosions of supernovae can produce a wealth of substructures in the diffuse hot gas and significantly affect the spectroscopic properties of the X-ray-emitting gas. The differential emission measure in the temperature space has a broad lognormal-like distribution. Such distribution enhances the X-ray emission at both low and high energy bands. We further show that the SN Ia ejecta is not well-mixed with the ambient medium and the X-ray emission is primarily from the shocked stellar wind materials which in general have low metallicities. These 3-dimensional effects provide a promising explanation to the above “missing feedback” and “missing metal” problems. In addition, we demonstrate that the supernova iron ejecta forms a very hot bubbles, which have relatively larger radial velocities driven by buoyancy, resulting in a smaller iron mass fraction in the bulk outflow. These distinct properties give a natural explanation to the observed positive iron abundance gradient which has been a puzzle for decades.