We report results from the analysis of XMM-Newton and INTEGRAL data of IGR J16479-4514. The unpublished XMM-Newton observation, performed in 2012, occurred during the source eclipse. No point-like X-ray emission was detected from the source, conversely extended X-ray emission was clearly detected up to a size distance compatible with a dust scattering halo produced by the source X-ray emission before being eclipsed by its companion donor star. The diffuse emission of the dust-scattering halo could be observed without any contamination from the central point X-ray source, compared to a previous XMM-Newton observation published in 2008. Our comprehensive analysis of the 2012 unpublished spectrum of the diffuse emission as well as of the 2008 re-analysed spectra extracted from three adjacent time intervals and different extraction regions (optimized for point-like and extended emission) allowed us to clearly disentangle the scattering halo spectrum from the residual point-like emission during the 2008 eclipse. Moreover, the point-like emission detected in 2008 could be separated into two components attributed to the direct emission from the source and to scattering in the stellar wind, respectively. From archival unpublished INTEGRAL data, we identified a very strong (3$times$10$^{-8}$ erg cm$^{-2}$ s$^{-1}$) and fast (25 minutes duration) flare which was classified as giant hard X-ray flare since the measured peak-luminosity is 7$times$10$^{37}$ erg s$^{-1}$. Giant X-ray flares from SFXTs are very rare, to date only one has been reported from a different source. We propose a physical scenario to explain the origin in the case of IGR J16479-4514.