My Django + Next.js site is containerised, lives on a single server, and everything looks secure, yet every few hours the public front-end dies with a 403 timeout. A quick docker-compose restart brings it straight back, but the cycle repeats. What I already know • I have read the logs and they report network-related errors just before the containers stop responding. • Custom firewall and network rules are in place, so something in that stack might be choking the traffic. • Nothing obvious appears in the application code itself. What I need from you 1. Trace the exact root cause—whether it’s Docker networking, nginx/Traefik, UFW/iptables, or something in the compose stack. 2. Patch or re-configure so the containers stay up indefinitely, without me having to babysit them. 3. Provide a concise summary of the fix (commands run, files changed) so I can reproduce it on staging. You will have SSH access to the server, docker-compose files, and the existing firewall ruleset. If extra monitoring or tweaks (e.g., health-checks, restart policies, rate limiting) are required, feel free to implement them as part of the solution. This is urgent—I’d like to stop restarting containers today—so please be prepared to dive in immediately and keep me posted as you progress.