Hi,
currently we have a Graphhopper server deployed which uses two storage volumes, one for the map and one for the cache (graph.location setting in the configuration YAML). While the Graphhopper is running and started with the map and cache from the two volumes, is it possible to start a second Graphhopper that acts on these volumes, i.e. it a) replaces the map in the map volume and b) deletes the cache in the cache volume and regenerates it, while the first one is running. While the second one is downloading the map and generating the cache, is it still possible to use the first Graphhopper? Does it load the map and cache in-memory, or is it required to have the files on disk?
My preliminary testing indicates that this should work, but is that stable?