Over the years, the number of servers I need to keep up with has increased. From just 1 in my house, to 3 or 4. Today, I need to keep up with 3 in my house (Consolidation will drop this down to just 2 soon), 2 on AWS, 1 in MD, and 1 in DE. Eventually the need to use Chrome remote desktop to get into the two off site servers got annoying. And the lack of lower interval data from the AWS lightsail admin pages made seeing what was happening difficult.
With Grafana, I have a one stop shop for server monitoring. And if I wanted it, alerting as well. For now, only the 5 Unraid hosts and the 2 Ubuntu based AWS servers have warranted installing telegraf. The few linux and Windows VM's running on those 5 UnRaid hosts are not at the point where monitoring is as important. Which is good, since there's a finite amount of computational power and memory on my two AWS VM's.
The end result of all of that, are the two dashboards you see below. Which are for the two servers I have running in AWS Lightsail. Meridian Gate handles all of the front end things like WordPress, static HTML, and Grafana. Stargate, is where all of my telegraf clients report to. It's only running InfluxDB. That segregation is one thing I'm very proud of since it keeps both of them responsive.
In addition to the above graph and alerting for the AWS server InfluxDB and Grafana run on, I have a fair number of others as well. Case in point, this dashboard, appropriately named "Detailed Dashboards" which I use to get a global view of everything on my UnRaid servers. On this page, you can quickly see that the server is largely idling, with a fair bit of things cached in RAM, and a constant but small amount of network traffic. That traffic is Telegraf. Although it is pretty interesting how quickly things change when Plex starts getting used.
