Live health & logs

Give a node a health URL and C4Stage's server probes it with a plain GET: 2xx is healthy, a 2.5 second timeout, redirects not followed, results cached for a few seconds, and the URL never sent to the browser. The canvas asks for the status every 30 seconds, and every 3 while a flow is in Live mode. A healthy node shows a quiet dot; a node that is down glows; a node with no health URL stays grey. A collapsed parent rolls up to the worst state anywhere beneath it, so problems surface even from five containers deep.

• Configure per node in the editor (Health url), or see every node's health and logs config at once in the editor's Monitoring tab.
• The Test button probes a URL server-side before you save it, and reports the HTTP status and the latency — or just Unreachable.
• Probes go to public addresses only: http and https, and never to localhost, a .local or .internal host, or an address that resolves into a private, loopback or link-local range. Such a URL is refused rather than probed.
• With Elasticsearch configured, the details sidebar's Latest logs shows a node's newest lines and offers Open in Kibana.
• Live health also drives flows in Live mode, the fastest way to see which requests an outage blocks. It is off on share links and in presentations; the public demo is the exception.