At present, in the Budker Institute of Nuclear Physics there is in operation the most complete set of modern mirror machines based on different principles of plasma confinement. This set includes the multi-mirror system GOL-3 for confinement of dense plasma heated by relativistic electron beam, the gas dynamic trap (GDT) for confinement of collisional plasma and anisotropic fast ions, and the ambipolar trap AMBAL-M.Mirrors have a number of advantages in comparison with the closed magnetic systems like tokamak, stellarator, etc. The most important are the following. The effects of disruptions are not appeared in mirrors. There are no divertor problems in them. Plasma pressure in a mirror device can be comparable with magnetic field pressure. As to the multi-mirror system, in this case, the β value can be even significantly higher than unity (the so called <<wall confinement>>). At last, mirrors are convenient for direct conversion of energy of charged particles leaving out the ends. This circumstance can turn out to be especially important in future <<low neutron>> schemes of fusion reactors.In principle, mirrors are very attractive from the engineering point of view, if the plasma confined in axisymmetric magnetic systems would be MHD stable. At present, the problem of MHD stability has already been solved for all axisymmetric traps designed in Novosibirsk. At least, the value β [approximately or equal to] 0.4 was obtained in these traps without any indications of macroscopic instability development.Some important results were obtained recently in the GOL-3 experiments. A specific mechanism of reduction of longitudinal electron thermal conductivity was detected. Recently, the magnetic system of the GOL-3 device was reconstructed into multi-mirror configuration. In this configuration, a new mechanism of fast ion heating was observed. As a result, the ion temperature was increased from a few eV up to 1 keV and the confinement time significantly increased (up to 1ms). These results were obtained for plasma density of 1021m-3.The experiments on the GDT device are directed to the solution of the problem of creation of a 14 MeV high power neutron source. At present, new powerful neutral beam injectors for this device are under construction. After installing these neutral beams at GDT, the electron temperature of plasma should increase up to 300 eV. It means that calculated neutron flux for the case of D-T reaction will be about 0.5 MW/m2. If this value will be obtained, it will be immediately possible to begin the design of final stage of high power neutron source.In the paper, the status of all mirror traps in Novosibirsk is presented and description of the main experiments is given.