This paper addresses data sustainability in smart agriculture, examining challenges and best practices throughout the agricultural data lifecycle. It differentiates between efficiency, sustainability, and resilience in data systems while analyzing key sustainability challenges including infrastructure limitations, data quality issues, and ownership concerns. The research proposes and details a conceptual framework for quantifying data sustainability, based on dimensions such as data relevance, utility, criticality, age, and cost. A Romanian agriculture illustrative application illustrates sustainability challenges across different farm scales with targeted improvement strategies. The findings contribute to understanding how sustainable data management can enhance agricultural practices while balancing technical, environmental, economic, and social dimensions for long-term viability.