A method for allocating a video conferencing task within a computer network comprising a plurality of candidate processing devices, the method comprising: for each of the candidate processing devices: determining a topological score, the topological score being a function of at least one end-to-end delay between a pair of participating endpoints through the candidate processing device; for each candidate processing device within at least a subset of the candidate processing devices: determining an availability score, the availability score being a function of an available processing capacity of the candidate processing device; selecting a preferred processing device among the candidate processing devices as a function of the topological scores and of the availability scores, and allocating the video conferencing task to the preferred processing device.