Basic Web

We use the Nagios check_http plugin, one built on curl (curl -k -i -s -m), or one built on the Python Requests library, to interact with URLs. A standard monitor retrieves a text string selection from the given URL within a specified warning/critical timeframe.

We need the full path and port, firewall access from our servers, and a text string returned with the page content. Some fields, especially those generated by javascript, may not be available to a basic web monitor.

Advanced Web

We may be able to use licensed emulation software to perform more complex interactions, but the results may be less transparent or predictable. If the application administrators can configure a local account with basic authentication, and limit the privileges to a designated action, we should be able to report a Pass/Fail on that action. Again, specific firewall privileges must be configured.

Our customer may write any script that returns output according to the Nagios Plugin Guidelines. Such a script may be hosted on the customer's server with the NCPA agent installed, in which case the script would be executed within the customer's environment. Nagios would, via the agent, have access only to the single execution command, and the script return code and string output. Nagios server agent reference: basics & links This allows for the calling of executable files without sharing the existence of those files outside the server: the process can be as complex as the customer is willing to maintain, as long as it yields a return code 0-3 ('OK':0, 'WARNING':1, 'CRITICAL':2, 'UNKNOWN':3).

Vendor Status Pages

Many cloud service vendors offer service status pages and/or RSS feeds. These may offer an account-specific API, or a service-wide overview of component Status. They may offer the service team the option of direct emails, or in some cases we may be able to monitor the API, feed, or page.