How to implement server monitoring on a system that prohibits outgoing network connections?

If outgoing network connections are prohibited on the server, monitoring should be organized so that the server itself sends status data (push model), rather than waiting for external requests.

Implementation options:

• Agent-based monitoring: an agent runs on the server, collecting metrics and periodically sending them to an external monitoring server via allowed channels (e.g., through an internal proxy or VPN).
• Local collection and upload: the server collects logs and metrics locally, and then the administrator periodically uploads them manually or through secure channels.
• Use of push-supported systems: for example, Prometheus Pushgateway, where the server sends data, and the monitoring system collects it.

If outgoing connections are completely prohibited, monitoring should be implemented within a network that has access to the server, for example, via SNMP or agent solutions operating within the local network.