Support for Drupal 7 is ending on 5 January 2025—it’s time to migrate to Drupal 10! Learn about the many benefits of Drupal 10 and find migration tools in our resource center.
To reproduce:
1. Choose any web site on D7 platform.
2. Caching options default - it does not matter if Redis is enabled or not, also problem stays with dev. alias
3. Put site into maintenance mode, take a look at front page as anonymous visitor
4. Put site live
For some time there is maintenance mode page shown. Seems like for 300 seconds as speedboster setting
Comments
Comment #1
omega8cc CreditAttribution: omega8cc commentedThis should be fixed in #1528996: Nginx microcaching should use fastcgi_cache_valid with 1s only for upstream errors.
Comment #2
omega8cc CreditAttribution: omega8cc commentedComment #3
omega8cc CreditAttribution: omega8cc commentedSo no, it doesn't help, because we force minimum TTL in the
global.inc
anyway.Comment #4
omega8cc CreditAttribution: omega8cc commentedWe need to reverse our TTL logic in the
global.inc
file to use Nginx default 1s infastcgi_cache_valid
instead of 300 as it is now. This may be tricky, but we need this to avoid caching not only 503 but also WSOD/500 for 5 minutes.Comment #5
omega8cc CreditAttribution: omega8cc commentedRelated follow-up in #1528996: Nginx microcaching should use fastcgi_cache_valid with 1s only for upstream errors.
Comment #6
omega8cc CreditAttribution: omega8cc commentedThis has been fixed in BOA head.
Comment #8
wprowe CreditAttribution: wprowe commentedRunning Drupal Core 7.26. I'm still seeing this problem. Using drush to perform site updates. URL is www.musicphotographers.net.