Hello everyone, we're planning to simplify the monitoring operations of systems which are running behind some kind of network barrier (e.g. firewall, NAT) and would like to hear what are your requirements and expectations. Prometheus architecture requires that all monitored targets are accessible from the server via TCP. The monitored data is most commonly collected via exporters which expose the measurements on dedicated ports. It implies that multiple ports have to be open on the firewall [1] which in some cases is not desired. To simplify the setup we're considering installing a reverse proxy on the managed system. We've got several options: 1. nginx or other general purpose web proxy server It is the most commonly used and most mature solution. The main concern we have here is that the users would possibly want to run nginx for own applications as well which would create a risk of breaking the configuration. 2. Exporter Exporter [2] It is a dedicated Prometheus reverse proxy. Ergo, the above problem is addressed. 3. PushProx [3] Follows a different design than the above two with proxy and client installed on both sides of the network barrier. It allows to keep all the configuration on the server side. The project is still work in progress but has a charm of not requiring any ingress ports to be open. I'd be happy to hear what are the expectations of the users and which of the described solutions would address your needs best. Thanks for your feedback Witek [1] https://www.uyuni-project.org/uyuni-docs/uyuni/administration/monitoring.htm... [2] https://github.com/QubitProducts/exporter_exporter [3] https://github.com/prometheus-community/PushProx -- To unsubscribe, e-mail: uyuni-users+unsubscribe@opensuse.org To contact the owner, e-mail: uyuni-users+owner@opensuse.org