Network management evolved in a way where implementing complex, high level network policies, implies dealing with some attributes that depend on low-level specific configuration. This reflects on a difficulty of changing the underlying infrastructure. SDN (Software-Defined Networking) concept opens a road for new developments due to the centralized non vendor-specific control of the network, most of it related with the separation of data and control planes. SDN raises different perspectives on how networks can operate and, consequently, how they can be managed and monitored. In particular, facing the undeniable need to reduce the amount of monitoring data in today's broadband networks, packet/flow sampling has emerged as one promising field for SDN. In this context, this paper is focused on exploring the SDN architecture, and its elements, for supporting sampling-based network monitoring. The aim is to take advantage of the integrated view of SDN controllers to apply and configure appropriate sampling techniques in network measurement points according to the requirements of specific measurement tasks. This will allow a flexible and service-oriented configuration of network monitoring, allowing also to improve the trade-off between accuracy and overhead of the monitoring process. To pursue this, in this paper relevant SDN elements will be examined, alongside with existing monitoring solutions in the SDN research area. The analysis of these solutions led to the proposal of a new approach for the flexible configuration of sampling-based monitoring resorting to SDN components and protocols. This flexibility also enables programmable measurements, allowing a SDN controller to manage measurement tasks concurrently at multiple spatial and temporal scales. Since collecting actual data to create information is important at the time of taking decisions, network operators need to understand the dynamic of their network through monitoring and sampling.