The STS literature has shown the diversity of openness practices; but little is known about how the HSS research communities are empirically dealing with the contemporary open data movement (ODM) in the "age of open science". In the context of the increasing digitization of practices and results, this movement, now transposed in European political agenda, has an impact on the very core of research activities and its support. Faced with new requirements of funding agencies, a growing number of studies from French academic institutions focus on data management at the laboratory and university levels, most often to support internal policies aiming at implementing data management tools. But how are qualitative research practitioners affected by the ODM? How do they perceive the injunctions and norms in favor of open data, and integrate openness in their research practices? What shifts does it bring about? The communication aims to present the empirical results of the PARDOQ project, which is the first qualitative survey conducted in France on the practices and representations of data sharing and data protection within research communities practicing or accompanying the collection of ethnographic materials. The method used relies on 19 semi-directed interviews with members of research support networks and with practitioners achieved in fall 2020, completed with a critical discourse analysis and participant observation. Results emphasize a considerable gap between ethnographers' practices and institutional expectations, both in terms of openness and compliance with the laws in force (security, GDPR). For those interviewed, the institutionalization of ODM is part of a broader context of scarcity and concentration of resources, and loses legitimacy because of the absence of sustainability of the latter. Sharing practices are logically built outside the ODM framework. Our research thus suggests that ethnography is a limit case that allows us to understand the limits of the ODM.