Domoticz

I found out today, after it failed for the first time after a long period, that the the NGINX config would only resolve the dns of api.honeywell.com one time, so it was pointing to a fixed ip address of a server that didn't work anymore. After some research I think the solution is to make these changes in the config of the reverse proxy to force the dns request each time:

jvdz wrote: ↑Friday 08 December 2023 14:52

Code: Select all

server  {
	include    /etc/nginx/proxy_params;
	listen 443 ssl;
	keepalive_timeout 70;
	server_name api.honeywell.com;
	## added this to force dns resolution  -> https://forum.nginx.org/read.php?2,215830,215832#msg-215832
	set $backend "api.honeywell.com";
	ssl on;
	ssl_certificate             /etc/ssl/letsencrypt/SSLcertificate.crt;
	ssl_certificate_key         /etc/ssl/letsencrypt/SSLprivatekey.key;
	access_log /var/log/nginx/honeywell.access.log;
	error_log /var/log/nginx/honeywell.error.log;
	resolver 8.8.8.8 valid=10s;
	location / {
		proxy_set_header Host api.honeywell.com;
		## Changed this to force dns resolution  -> https://forum.nginx.org/read.php?2,215830,215832#msg-215832
		# proxy_pass https://api.honeywell.com;
		proxy_pass https://$backend;
		sub_filter '"expires_in":"1799"' '"expires_in":"699"';
		sub_filter 'Invalid' 'foutje';
		sub_filter_types *;
	}
}