Strange behavior of nginx + php7.0-fpm on Ubuntu 16.04. What could it be?

Question

PHP

Oddities are manifested in the following:
- the site works for some time, the time is always different, it works quite quickly and without problems
- after some time *26481 recv() failed (104: Connection reset by peer) while reading response header from upstream by logs Nginxa
- reboot php-fpm and everything starts working fine again until next time
host configs in nginx

server {
  listen 80;
  server_name ;
  return 301 https://$server_name$request_uri;
}
 server {
        charset utf-8;
        client_max_body_size 128M;

        autoindex on;
  server_tokens off;

        listen 443 ssl;

  ssl_session_timeout 1d;
  ssl_session_cache shared:SSL:50m;
  ssl_session_tickets off;	
  ssl_protocols TLSv1.2;
  ssl_ciphers 'ECDHE-ECDSA-AES256-GCM-SHA384:ECDHE-RSA-AES256-GCM-SHA384:ECDHE-ECDSA-CHACHA20-POLY1305:ECDHE-RSA-CHACHA20-POLY1305:ECDHE-ECDSA-AES128-GCM-SHA256:ECDHE-RSA-AES128-GCM-SHA256:ECDHE-ECDSA-AES256-SHA384:ECDHE-RSA-AES256-SHA384:ECDHE-ECDSA-AES128-SHA256:ECDHE-RSA-AES128-SHA256';
  ssl_prefer_server_ciphers on;
  add_header Strict-Transport-Security "max-age=15768000";
  ssl_stapling on;
  ssl_stapling_verify on;	

  location / {
          try_files $uri $uri/ /index.php?$args;
                proxy_connect_timeout 300;
                proxy_send_timeout   600;
                proxy_read_timeout   600;
                proxy_buffer_size    64k;
                proxy_buffers     16 32k;
                proxy_busy_buffers_size 64k;
                proxy_temp_file_write_size 64k;
                proxy_pass_header Set-Cookie;
                proxy_redirect     off;
                proxy_hide_header  Vary;
                proxy_set_header   Accept-Encoding '';
                proxy_ignore_headers Cache-Control Expires;
                proxy_set_header   Referer $http_referer;
                proxy_set_header   Host   $host;
                proxy_set_header   Cookie $http_cookie;
                proxy_set_header   X-Real-IP  $remote_addr;
                proxy_set_header X-Forwarded-Host $host;
                proxy_set_header X-Forwarded-Server $host;
                proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
        }

        location ~ \.php$ {
                include snippets/fastcgi-php.conf;
                #fastcgi_pass unix:/run/php/php7.0-fpm.sock;
    fastcgi_pass 127.0.0.1:9000;
    fastcgi_read_timeout 300;
        }

  #ENABLE GZIP
  gzip on;
  gzip_disable "msie6";
  gzip_types text/plain text/css application/json application/x-javascript text/xml application/xml application/xml+rss text/javascript application/javascript;
  gzip_comp_level 5;

  #ENABLE BROWSER CACHE
  location ~* ^.+\.(woff|rss|atom|jpg|jpeg|gif|png|ico|rtf)$ {
    expires max;
  }

  location ~ /.well-known {
                allow all;
        }
}

nothing changed in the fpm/www.conf settings except

;listen = /run/php/php7.0-fpm.sock
listen = 127.0.0.1:9000

in the slow fpm log after the crash writes index.php, the one that is called by the nginxa virtual host.
I tried tuning fpm and timeouts. Didn't give any result.
Who can faced such as won? help me please

Answer 1

the question was closed. found the shit code of one of the developers, which looped the backend, as a result of which there was no exit from the procedure and fpm did not understand what to do. thank you all for your attention

Answer 2

There was such a problem when FPM lay down randomly on a VPS with Debian. At the same time, the number of processes and so on did not help. Do not believe me - I removed xDebug and there were no more problems. But this is in my case.