Scaling NABox or NE Harvest 2.0

[coreymr]

Hi, in parallel to my other inquiry I was wondering if you had a reference as to how well either NABox or Harvest 2.0 scales? We are a geo-dispersed org with 130+ filers in multiple regions/countries with varying degrees of connectivity.

Our OCUMs are setup with regional instances. Should we think about setting up regional NAbox as well to limit "overloading" the Harvest/Grafana functions? Or should be okay using one "pane of glass" for our environment?

[rahulguptajss]

1. one data point for Harvest 2.0 - there are customers monitoring 300 nodes on a single instance of Harvest running on a machine with 24 GB of memory. You can try scaling it further and see.
2. @ybizeul can provide more information on NABox