Nagios XI scalability

[skynardo]

We are considering purchasing Nagios XI and are curious as to the maximum number of Nodes and services that can be (or are being) monitored on a single monitoring server. If we bought the unlimited enterprise edition license and ran this on a RHEL 6 12 CPU blade with 64Gb of ram, what would the aproximate limit be to our unlimited license? We would be monitoring Windows, Linux and AIX servers, lets say 10 services per machine using ncpa/nrpe. We would want to collect historical metrics minimally for all production Nodes. The hardware requirement doc has a bottom row indicating >500 Nodes >2500 services. Im looking for the < side of that estimate and or accounts of actual customers and what they are able to do on a single beefy monitoring server. Thanks

[slansing]

You should be well within the means of that server with 500 hosts, and roughly 2500 services. Unfortunately, it is really hard to give a ballpark estimate since everyone checks systems differently. For instance, active SNMP checks can be quite CPU intensive, especially when you go to sub 5 minute check intervals, where-as passively received NRDS results, don't really make a dent at all on the XI server's end. Depending on how you plan to be checking these systems (which what protocols) and how often you plan to check them on average (both hosts and services) we would be able to offer a bit of a better opinion. I would say though that 500 hosts, 2500 services with a mix of active and passive checks should be doable just fine on that server.