Vacuum breakdown, also known as vacuum discharge, is a common phenomenon in nature and gains an increasingly important role in modern technologies. In spite of a remarkable advance in understanding of the nature of the breakdown, the role of an anode, i.e. a positively charged electrode, in the development of the breakdown is still completely unclear. In this paper, we employ a streak camera with picosecond time resolution to observe precisely the evolution of anodic glow from different anode materials. The results show that the choice of the anode material does not affect neither the delay time between the cathodic and anodic flares, nor the formation of the conductive channel. Furthermore, we show that the heating of the anode surface by runaway electron currents is not sufficient to evaporate enough atoms for the anodic glow. On the other hand, we show that the neutrals for the anodic flare can be produced by the ions from the expanding cathode plasma by sputtering. Finally, the coincidence in time of the voltage collapse and the anode glow is consistent with the fast expansion of the cathode plasma, which causes both the voltage collapse and the anode glow when it reaches the anode and densifies by sputtering and reflection. However, the two events are not in direct dependence of one another, since the order of the appearance of them is random, implying that a fully conductive channel can be established without any light emission from the anode.