The popularization of cloud computing has provided the emergence of large volumes of data that are stored in Data Centers (DCs). These locations store data of different types, origins, and priorities for their owners. The DCs are subject to natural or man-made attacks. The attacks are diverse and happen quickly after their detection. Therefore, this paper proposes two techniques to evacuate data from threatened DCs to those who are outside the risk zone of the attack. The first technique is based on the Service Level Agreement (SLA) of the data and the second one is based on the order that they arrive at the DC, using the algorithm LIFO. Both techniques performed similarly on the amount of evacuated data and at the time for evacuation. However, the SLA policy distributes the data on a priority scale according to the SLA, while the LIFO policy ranks data on the same scale's priority.