Single-greatest runup measurement was 34.9 meters!
http://pmel.noaa.gov/tsunami/indo20041226/itic_team.htm
That height was in a valley, but numerous other heights were in the 30-meter range on the west coast. This range seems inordinately high for a far-field quake-generated event, even given the unusual topography & seafloor contours in that area (unusual contours often funnel wave energy into a concentrated area). What I'm suggesting by implication is that the main quake might have triggered one or more near-field submarine slides or slumps, much the way that slides from quakes in 1946 (Alaska & Hawaii), 1964 (Alaska) & 1998 (Papua New Guinea) generated waves much higher than would naturally be expected from the quake faults themselves. (Typically, a seabed vertical thrust of 15 meters would create a tsunami of no more than 15 meters. The displacement from the Sumatran quake, in & of itself, probably shouldn't have created a wave with a 34-meter runup.)
("Runup" is maximum vertical height onshore above sea level.)