I am using tileserver-gl-light to serve vector tiles. Currently I'm load testing the server, at specific values there is a drastic increase in request latency. However the CPU performance of the Virtual Machine - which I am using as server - stays below 20%. Does this mean I could decrease the latency by allocating more memory to the tileserver-gl-light proces, or does that have nothing to do with it?
0 Answers
Related