We study the orbital architecture, physical characteristics of planets, formation and long-term evolution of the Kepler-30 planetary system, detected and announced in 2012 by the KEPLER team. We show that the Kepler-30 system belongs to a particular class of very compact and quasi-resonant, yet long-term stable planetary systems. We re-analyse the light curves of the host star spanning Q1-Q17 quarters of the KEPLER mission. A huge variability of the Transit Timing Variations (TTV) exceeding 2 days is induced by a massive Jovian planet located between two Neptune-like companions. The innermost pair is near to the 2:1 mean motion resonance (MMR), and the outermost pair is close to higher order MMRs, such as 17:7 and 7:3. Our re-analysis of photometric data allows us to constrain, better than before, the orbital elements, planets' radii and masses, which are 9.2 ± 0.1, 536 ± 5, and 23.7 ± 1.3 Earth masses for Kepler-30b, Kepler-30c and Kepler-30d, respectively. The masses of the inner planets are determined within ∼ 1% uncertainty. We infer the internal structures of the Kepler-30 planets and their bulk densities in a wide range from (0.19 ± 0.01) g·cm −3 for Kepler-30d, (0.96 ± 0.15) g·cm −3 for Kepler-30b, to (1.71 ± 0.13) g·cm −3 for the Jovian planet Kepler-30c. We attempt to explain the origin of this unique planetary system and a deviation of the orbits from exact MMRs through the planetary migration scenario. We anticipate that the Jupiter-like planet plays an important role in determining the present dynamical state of this system. tems make it possible to further refine their orbital parameters and masses (e.g., Borsato et al. 2014;Barros et al. 2014).In this paper, we report on a new and up-to-date characterization of the Kepler-30 system composed of three planets. These planets have been validated by Fabrycky et al. (2012) on the basis of the very initial Q1-Q6 KEPLER data quarters ( 600 days) as well as by Tingley et al. (2011) on the basis of the first Q0-Q2 quarters.The early studies on this system were focused on the orbital-spin alignment and star-spots detection (Sanchis-Ojeda et al. 2012) or on differential rotation of the parent star (Lanza et al. 2014). Here, we take full advantage of the available light curves of Kepler-30, spanning Q1-Q17 quarters ( 1600 days) to constrain better the masses and orbital elements of the planets.We find the Kepler-30 system particularly interesting and unique in the KEPLER sample for several reasons. One of them is the huge TTV full-amplitude of the innermost planet of 2 days (48 hours), which is two times larger than the TTV fullamplitude of 1 day for the innermost planet in the Kepler-88 system (Nesvorný et al. 2013) dubbed as "the king of the transit c 2017 RAS