The resilience of an active distribution network (ADN) is continuously threatened by extreme weather, natural disaster, and cyber-physical incidents. An automatic restoration framework is thus desired to optimally integrate the emerging smart-grid technologies with the grid-edge flexibility to self-heal the ADNs. This paper proposes a framework for sequential ADN restoration by coordinating different distributed energy resources (DERs) on the grid-edge. The uncertainty of the renewable DERs is, however, a challenge. Therefore, the proposed framework is formulated as a chance-constrained optimization problem (CCOP). The chance-constraints use a parametrized response policy for the control during fluctuations and are reformulated for an efficient solution. The proposed CCOP's solution provides a sequence of control actions at each time-step that coordinates the network switches and DERs to form multiple islanded microgrids. Its effectiveness is evaluated on IEEE 123 node ADN.