There was a time in satellite remote sensing history when a laboratory could processes all of the recorded data from the satellites that it tracked on one computer and most likely have time to spare for doing other tasks within the laboratory. Fortunately computers and disk storage technology has increased faster than the satellite sensor technologies. So, even today one computer can process all recorded data from any one satellite and its sensors. However, there are many more satellites, and many more sensors in orbit today creating a need for scalable processing. In other to achieve the above performances, satellite data processing at Naval Research Laboratory Marine Meteorology Division (NRLMRY) switched from single computer per satellite, to a Linux cluster of one master, and eleven nodes. Upon reception of a data stream, the product generations are queued to the PBS queuing system for production of the products made from this data stream. Present work presents the short- and long-term strategy adopted for keeping pace with the ever-growing need for satellite products. NRLMRY currently processes 22 sensors from 26 satellites globally, producing over 35,000 image products daily in an automated fashion. These images can be mostly divided into three different earth surface resolution sizes: i) Global and full disk imagery; ii) Mesoscale imagery, and iii) High resolution (1 km or less) imagery. In addition to satellite imagery there are many other value added products, such as overlaid imagery of numerical model output and/or derived satellite products such as cloud top properties or rain. The focus of all these efforts is to produce web-based quick-looks for operational use by general marine meteorology community.