I am the complete IT department in a small company, not so many servers to monitor, and to keep it manageable alone I monitor endpoints too, as the article linked by @Norm calls them, that are inside company network.
Our endpoints have some of the issues and solutions implemented that are mentioned in that article.
Not covered in the article is that wired endpoints that going up and down, will cause network interface services on the connected switch to send notifications too. So currently I have those ‘silenced’.
Endpoints outside the company are not checked yet. Need to change VPN infrastructure to make those accessible, like ping and more. It’s on my nice to have list in the future.
Laptops inside the company are default all year in downtime.
For the desktops I have check periods, based on office hours, and scheduled downtime based on personnel roster. Since I am using RAW, scheduled downtime is done with a workaround, which I explain in this article: RAW: Recurring scheduled downtimes workaround using rest api, bash & cron
With (un)planned absence of personnel I still have to manually add their endpoints in downtime.
I still need to look into making scripts on endpoints that can run when they boot or shutdown.
Notifications of endpoints are send to a different mailbox, which is used as /dev/null .
Still need to figure out to not send those notifications at all or for just some of their services.
I mostly monitor endpoints to see if they are up, system time, need system updates. I am working towards replacing the current tool PDQ Inventory (Free Mode version) with Checkmk RAW.
I have endpoints and servers in a Nagvis map of the office, so I have a quick overview of their status. I use the Nagvis map also for keeping track of extra unmonitored information too, like placement of network sockets, default assigned workspace to personnel etc. To avoid turning the map into a jungle of information I turn on/off certain layers explained here: NagVis: how to turn on/off layer(s) with info text boxes
Example for why I monitor endpoints: we use VOIP desktop phones and they currently have an issue that they sometimes loose there assigned IP address and are no longer giving service. With Checkmk and the Nagvis map I see quickly when the issue arise and fix it.