If you see this page, the nginx web server is successfully installed and working. Further configuration is required.
For online documentation and support please refer to
nginx.org.
Commercial support is available at
nginx.com.
The node serving this page (edge-ams-12.glaciernet.example) operates as a transparent caching proxy for a small set of internal services. HTTP requests directed at this hostname that do not match a known route are returned this default page rather than a 404, because some health-check clients in our environment treat any non-2xx response as an outage and would otherwise generate noisy alerts during routine probes.
Operators wishing to verify reachability should use the /_status endpoint instead. That endpoint is restricted by network ACL and is not visible on the public internet. There is no shell access, FTP service, or administrative console exposed by this host. Edge configuration is replicated from the control plane every two minutes and any drift is reconciled automatically without operator intervention.
The cache layer stores responses for fifteen minutes by default. Backend services may override this by sending an explicit Cache-Control header in the response. Stale-while-revalidate is enabled for read-only endpoints and allows the cache to keep serving the previous response while a fresh copy is fetched in the background. This significantly reduces the impact of short upstream outages and keeps tail latency under control during peak hours. Cache keys include the request method, the normalised path, and a small set of selected query parameters chosen by the route definition.
Per-route invalidation can be triggered from the operations dashboard. Manual cache flushes are recorded in the audit log along with the operator identity and the reason supplied at flush time. The audit log is shipped off the node every five minutes and retained for ninety days in the central archive. Bulk invalidations are deliberately throttled to prevent upstream services from being overwhelmed by the resulting refill traffic.
This edge accepts both HTTP/1.1 and HTTP/2 over TLS 1.2 and 1.3. Older TLS versions are not enabled. Each client IP is limited to a generous but finite number of concurrent streams; clients that exceed the limit will observe back-pressure in the form of delayed acknowledgements rather than hard rejections. Idle keepalive connections are held open for sixty seconds before being closed, which keeps the connection table compact while still allowing well-behaved clients to reuse sockets.
For any operational questions concerning this edge tier, please write to edge-ops@glaciernet.example. Routine incidents are tracked through the dedicated ticketing system; this mailbox is used only for non-urgent correspondence such as scheduled maintenance enquiries or peering proposals.
Thank you for using nginx.
nginx/1.27.0 — release channel: mainline — edge-ams-12