nominatim: clamp down on rate limits

[chef.git] / cookbooks / nominatim / templates / default / nginx.erb

diff --git a/cookbooks/nominatim/templates/default/nginx.erb b/cookbooks/nominatim/templates/default/nginx.erb
index 998fcdfaa5f24606d0c81ad5174bc6279a417f3d..af7ea960b3cc3799e17d25593dc942dfcd1a7915 100644 (file)
--- a/cookbooks/nominatim/templates/default/nginx.erb
+++ b/cookbooks/nominatim/templates/default/nginx.erb
@@ -1,11 +1,11 @@
 upstream nominatim_service {
 upstream nominatim_service {

-  server unix:/run/php/nominatim.openstreetmap.org.sock;
+  server unix:/run/gunicorn-nominatim.openstreetmap.org.sock fail_timeout=0;

 }
 
 map $uri $nominatim_script_name {
 }
 
 map $uri $nominatim_script_name {

-    ~^(.+?\.php)         $1;
-    ~^/([^/]+)           $1.php;
-    ^$                   search.php;
+    ~^/*(.+?)\.php        $1;
+    ~^/*([^/]+)           $1;
+    ^$                   search;

 }
 
 map $uri $nominatim_path_info {
 }
 
 map $uri $nominatim_path_info {


@@ -60,9 +60,20 @@ geo $whitelisted {
     8.43.85.3 1; # gnome
     8.43.85.4 1; # gnome
     8.43.85.5 1; # gnome
     8.43.85.3 1; # gnome
     8.43.85.4 1; # gnome
     8.43.85.5 1; # gnome

+    2620:52:3:1:5054:ff:fe0a:75a4 1; # gnome
+    2620:52:3:1:5054:ff:fe0a:75a2 1; # gnome
+    2620:52:3:1:5054:ff:fe0a:75aa 1; # gnome
+    34.234.151.67 1; # gnome - https://github.com/openstreetmap/operations/issues/1160
+    54.165.53.199 1; # gnome - https://github.com/openstreetmap/operations/issues/1160
+    35.153.15.118 1; # gnome - https://github.com/openstreetmap/operations/issues/1160

 }
 
 }
 

-map $missing_email$missing_referer$http_user_agent $blocked_user_agent {
+map $server_protocol$http_user_agent $cleaned_user_agent {
+    default $http_user_agent;
+    "~^HTTP/1..Mozilla/" Script$http_user_agent;
+}
+
+map $missing_email$missing_referer$cleaned_user_agent $blocked_user_agent {

    default 0;
    "11" 2; # block any requests without identifier
    include <%= @confdir %>/nginx_blocked_user_agent.conf;
    default 0;
    "11" 2; # block any requests without identifier
    include <%= @confdir %>/nginx_blocked_user_agent.conf;


@@ -101,7 +112,7 @@ map $whitelisted$generic_mozilla$uri $limit_reverse {
     ~02/reverse.*  $binary_remote_addr;
 }
 
     ~02/reverse.*  $binary_remote_addr;
 }
 

-limit_req_zone $limit_www zone=www:50m rate=2r/s;
+limit_req_zone $limit_www zone=www:50m rate=1r/s;

 limit_req_zone $limit_tarpit zone=tarpit:10m rate=1r/s;
 limit_req_zone $binary_remote_addr zone=blocked:10m rate=20r/m;
 limit_req_zone $limit_reverse zone=reverse:10m rate=10r/m;
 limit_req_zone $limit_tarpit zone=tarpit:10m rate=1r/s;
 limit_req_zone $binary_remote_addr zone=blocked:10m rate=20r/m;
 limit_req_zone $limit_reverse zone=reverse:10m rate=10r/m;


@@ -138,8 +149,8 @@ server {
     ssl_certificate /etc/ssl/certs/<%= node[:fqdn] %>.pem;
     ssl_certificate_key /etc/ssl/private/<%= node[:fqdn] %>.key;
 
     ssl_certificate /etc/ssl/certs/<%= node[:fqdn] %>.pem;
     ssl_certificate_key /etc/ssl/private/<%= node[:fqdn] %>.key;
 

-    root <%= @directory %>/website;
-    index search.php;
+    root <%= @directory %>/static-website;
+    index /search;

 
     access_log <%= node[:nominatim][:logdir] %>/nominatim.openstreetmap.org-access.log combined;
     error_log <%= node[:nominatim][:logdir] %>/nominatim.openstreetmap.org-error.log;
 
     access_log <%= node[:nominatim][:logdir] %>/nominatim.openstreetmap.org-access.log combined;
     error_log <%= node[:nominatim][:logdir] %>/nominatim.openstreetmap.org-error.log;


@@ -175,48 +186,43 @@ server {
         add_header Access-Control-Allow-Origin "*" always;
     }
 
         add_header Access-Control-Allow-Origin "*" always;
     }
 

-    location @php {
-        if ($blocked_user_agent ~ ^2$)
-        { return 403; }
-        if ($blocked_referrer)
-        { return 403; }
-        if ($blocked_email)
-        { return 403; }
-        include <%= @confdir %>/nginx_blocked_generic.conf;
+    location ~* ^/(search|reverse)(\.php)?/ {
+        error_page 404 /404-old-search-syntax.html;
+        return 404;
+    }

 
 

-        limit_req zone=www burst=10;
-        limit_req zone=tarpit burst=5;
-        limit_req zone=reverse burst=5;
-        limit_req_status 429;
-        fastcgi_pass nominatim_service;
-        include fastcgi_params;
-        fastcgi_param QUERY_STRING    $args;
-        fastcgi_param PATH_INFO       "$nominatim_path_info";
-        fastcgi_param SCRIPT_FILENAME  "$document_root/$nominatim_script_name";
+    location @php {

         if ($forward_to_ui) {
             rewrite ^(/[^/]*) https://$host/ui$1.html redirect;
         }
         if ($forward_to_ui) {
             rewrite ^(/[^/]*) https://$host/ui$1.html redirect;
         }

-    }
-
-    location ~* \.php$ {

         if ($blocked_user_agent ~ ^2$)
         { return 403; }
         if ($blocked_referrer)
         { return 403; }
         if ($blocked_email)
         { return 403; }
         if ($blocked_user_agent ~ ^2$)
         { return 403; }
         if ($blocked_referrer)
         { return 403; }
         if ($blocked_email)
         { return 403; }

+        if ($args ~* "q=[0-9]+\.[0-9]+\.[0-9]+\.[0-9]+[ &]")
+        { return 418; }

         include <%= @confdir %>/nginx_blocked_generic.conf;
 
         limit_req zone=www burst=10;
         include <%= @confdir %>/nginx_blocked_generic.conf;
 
         limit_req zone=www burst=10;

-        limit_req zone=tarpit burst=2;
+        limit_req zone=tarpit burst=5;

         limit_req zone=reverse burst=5;
         limit_req_status 429;
         limit_req zone=reverse burst=5;
         limit_req_status 429;

-        fastcgi_pass    nominatim_service;
-        include         fastcgi_params;
-        fastcgi_param   SCRIPT_FILENAME    $document_root$fastcgi_script_name;

 
 

-        if ($forward_to_ui) {
-            rewrite (.*).php https://$host/ui$1.html redirect;
+        if ($request_method = 'OPTIONS') {
+          add_header 'Content-Type' 'text/plain; charset=UTF-8';
+          add_header 'Content-Length' 0;
+          add_header Access-Control-Allow-Origin "*";
+          add_header Access-Control-Allow-Methods 'GET,OPTIONS';
+          add_header Access-Control-Allow-Headers $http_access_control_request_headers;
+          return 204;

         }
         }

+
+        proxy_set_header Host $http_host;
+        proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
+        proxy_set_header X-Forwarded-Proto $scheme;
+        proxy_redirect off;
+        proxy_pass http://nominatim_service;

     }
 }
     }
 }