I am using a 12 node VM cluster . My Index size is about 200GB and currently I have 10 shards with 1 replica on a local gateway. I am also using 1 VM just as master with no data for coordination. I have 16GB RAM available per VM.I would like to improve performance of my ES so that I can serve about 25M queries a day. Currently I am using indices.memory.index_buffer_size: 8gb . Would increasing this help ?
Can anyone recommend a setup that would give me a quick performance boost.