Troubleshooting
The failures worth knowing by heart, and what each one means.
A white page or a 500 error
Read storage/logs/laravel.log first — it names the actual exception.
The two most common causes on a fresh server:
- Wrong database credentials. The error mentions
SQLSTATE[HY000] [1045]. - Cached files owned by another user. A cache built as
rootis unreadable byphp-fpmrunning aswww-data. Rebuild as the web server user:
sudo -u www-data php artisan config:cache
sudo -u www-data php artisan route:cache
sudo -u www-data php artisan view:cache
The site loads but has no styles
public/build is missing or the upload was interrupted. Re-unpack the release archive; the assets ship pre-built and the manifest must match the files.
"mysqldump: command not found" during a backup
The web server has a different PATH. Set the directory in .env:
DB_DUMP_BINARY_PATH=/usr/bin
An update fails with "No update-manifest.json"
The package is a full release zip, not an update package. Full zips are for a fresh install; updates are applied from the signed *-update-*.zip.
An update fails with "already exists"
A previous attempt was rolled back but the restore did not remove tables the package had created. Restore the pre-update backup from Admin → Backups, then apply the update again. The current release wipes the schema before importing, which removes this class of failure; older builds did not.
Uploads fail with a 413
nginx rejects the body before PHP sees it:
client_max_body_size 64M;
Then reload nginx.
Images upload but do not appear
php artisan storage:link was not run, or public/storage points at the wrong place. Remove the link and create it again:
rm public/storage
php artisan storage:link
The theme is active but the storefront is unchanged
Fixed in 1.2.0: on newer Laravel versions the view finder is not a singleton, so a theme could register its location in an object that was thrown away. Update to 1.2.0 or newer; re-activating the theme applies the location immediately.
API returns 500 with a rate limiter error
Routes were cached without the limiters being registered. Fixed in 1.1.1. On older builds run php artisan route:clear and do not cache routes.
License activation fails
See license activation — the checklist there covers the API URL, outbound access and the meaning of domain_conflict.
Everything is slow under load
In this order: enable OPcache, switch cache and sessions to a real cache store, then look at the database. A venue selling thousands of tickets a month runs comfortably on 2 GB; a slow page is nearly always an unindexed query or a missing cache, not the server.
Getting help
Write to support@scenecms.com with the version from Admin → System, the last lines of storage/logs/laravel.log, and what you did right before it broke. That is enough to answer most tickets in one reply.
Not covered here?
Write to sales@scenecms.com — we answer within one business day.