I’m pretty sure this isn’t a “supported” setup by Webflow, but I’m looking for advice. I’m seeing a strange behavior with excessive DNS caching in nginx while proxying to proxy-ssl.webflow.com, even though from what I’ve gathered my nginx configuration should be correct.
My configuration is:
nginx: 1.14.1
server { ... resolver 172.16.0.2 valid=10s; # Amazon AWS resolver location /elsewhere { other configurations; } # everything that's not already handled, proxy to webflow location / { # webflow.mydomain.com proxies to proxy-ssl.webflow.com set $webflowUpstream https://webflow.mydomain.com; proxy_pass $webflowUpstream; proxy_set_header Host webflow.mydomain.com; proxy_ssl_name webflow.mydomain.com; proxy_ssl_server_name on; proxy_next_upstream error timeout invalid_header http_500 http_502 http_503 http_504; proxy_cache_valid 200 5m; proxy_cache_valid 301 5m; proxy_cache_valid 302 5m; proxy_cache_valid any 1m; proxy_connect_timeout 5s; proxy_send_timeout 60s; proxy_read_timeout 60s; proxy_cache brochure-cache; proxy_cache_use_stale error timeout invalid_header updating http_500 http_502 http_503 http_504; proxy_cache_revalidate on; proxy_set_header X-Real-IP $remote_addr; proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for; proxy_set_header X-Forwarded-Proto $scheme; proxy_pass_header Referer; proxy_pass_header User-Agent; client_max_body_size 1024m; proxy_redirect https://webflow.mydomain.com/ https://mydomain.com/; }
If it’s not obvious from the configuration, we have other paths that proxy to other parts of our site (like our main application), so we can’t just configure our root domain to use Webflow’s DNS IPs – we have to proxy.
This config works great for a day or two, but then when proxy-ssl.webflow.com’s IPs change, nginx hangs on to the old IPs even though every nginx forum I’ve found shows this configuration should work – using a resolver and a variable for the proxy_pass.
I’m curious if anyone here has run into this issue. Any help anyone can offer would be appreciated.
Thanks!