Latest revision as of 02:00, 8 May 2026

Add these to settings.php:

 $settings['cache']['bins']['render'] = 'cache.backend.database'
 $settings['cache']['bins']['dynamic_page_cache'] = 'cache.backend.database'
 $settings['cache']['bins']['page'] = 'cache.backend.database'

reconfigure /etc/mysql/conf.d/docker.conf

 [mysqld]
 # skip-host-cache
 # skip-name-resolve
 max_connections = 300

part IV

See how long the request takes:

 curl -o /dev/null -s \
   -w 'dns=%{time_namelookup} connect=%{time_connect} tls=%{time_appconnect} ttfb=%{time_starttransfer} total=%{time_total}\n' \
   https://piousbox.com/issues/2025q2-issue

Add nginx log temporarily:

 log_format timing '$remote_addr - $request '
                 'rt=$request_time '
                 'urt=$upstream_response_time '
                 'uht=$upstream_header_time '
                 'uct=$upstream_connect_time';
 #
 access_log /var/log/nginx/timing.log timing;

Interpretation:

• high request_time + high upstream_*
  • → backend app is slow
• high request_time + no upstream times
  • → nginx/static filesystem issue
• high upstream_connect_time
  • → nginx struggling to reach upstream
• high upstream_header_time
  • → app slow before first byte

For containerized Drupal, a 3–4 second first-hit delay is commonly one of:

• PHP-FPM worker cold start
• Drupal cache rebuild/bootstrap
• slow DB query
• uncached Twig/container compilation
• network latency to DB/Redis
• bind-mounted filesystem slowness
• OPCache not configured
• reverse DNS / external HTTP calls

check php cache

 php -i | grep opcache

changes I'm making

 opcache.validate_timestamps=0 # made no difference
 realpath_cache_ttl = 600