In a wireless sensor network (WSN), different types of nodes (i.e., active, misbehaving, dead) may coexist. To ensure seamless performance of the network, network administrators often use several mechanisms to recover misbehaving or faulty nodes. As it is impossible for a node to predict the state of another node, the recovery attempts may not be successful every time. Such unsuccessful attempts may bring detrimental effects to the network. In this paper, we introduce three of such problems which occur due to change in state (such as dumb to dead, etc.) with such attempts. We name the problems as Recovery Attempt for Dead (RAD) node, Insecure Recovery Attempt (IRA), and Data Burst (DB). The detrimental effects of these problems are evaluated using network simulator 3 (ns3) simulations. Additionally, this paper also suggests some solution guidelines for these problems. The analysis and simulation experiments led us to enlist some open research issues in the concluding section.