Quick start monitoring by using lightweight monitors

Quick start monitoring by using lightweight monitors

Basic setup: use simple monitors

Biker race

There are instructions to get started with IPHost Network Monitor, covering most typical situations. However, using the defaults (for both monitors parameters and for network discovery) when looking for hosts and/or monitors can result in rather slow setup. In most cases, it is simpler to create “draft”, lightweight monitors just to make sure the service exists and accepts connections, and add actual protocol-specific checks as required. We provide sample instructions below: follow them to create and start your monitoring setup.

The initial monitors set can be added both manually and by running a network discovery; after that, the typical monitors sets can be copied over multiple hosts.

Use PING and TCP monitors only

The basic monitors include PING, TCP and UDP monitors. Unless a response is expected, using UDP doesn’t make much sense, since it won’t actually check the corresponding service presence.

PING monitor reports general device presence; keep in mind that certain services may block ICMP protocol (including PING), thus TCP may remain the only tool to check. An example of such service includes AWS Lightsail. Note that under normal circumstances ICMP should not be blocked.

TCP is a fundamental protocol standard, ensuring data delivery (reliable connections). TCP monitor checks for connection possibility only; if more complicated checks are expected, other monitor types should be used. This type of monitor may be used as dependency monitor: if certain TCP port must be open under normal circumstances, corresponding TCP monitor state can be used to prevent other monitors from running in case the port is closed.

The above two monitor types are suitable to create “quick” monitoring setup; add detailed check afterwards, as you see fit.

Network discovery

To speed up creating simple connectivity monitors, network discovery can be used, for both one-time lookup, and for periodic (scheduled) hosts list updating.

To look for several possible ports, at “Select monitor types” step of network discovery wizard, click “Ports” field for TCP monitor and enter comma-separated list of ports to scan for.

To map TCP/UDP ports to corresponding well-known services, you can use IANA reference. There’s also a short reference in so called “services” file, located in “/etc” in Unix-like and macOS operating systems, and in “C:\Windows\System32\drivers\etc” in Windows.

Cloning hosts and monitors

Another strategy is this: when there are several hosts using typical set of monitors, it could be quicker to create required set of monitors for a single host, and then either

  • copy the whole host (along with all its monitors) or selected monitors, as explained in Tree view manual, or
  • create an application template and assign it to other hosts, thus copying only the required monitors to other hosts

Note: creating template may be more flexible approach, since you can replace template afterwards, whenever required.

Conclusion

We recommend starting with simple monitors set, to create monitoring setups quicker, with lesser load on the system running IPHost. In case you would need advice on this, feel free to contact us, or leave us a comment below.