Working with elasticsearch cluster

Doesn't that make your collector a single-point-of-failure?