Update docker-compose.yml

[netherwraith]

After pulling the latest commits, loops-server-horizon and loops-server-loops couldn't be started again.

The log told

ℹ️ NOTICE (generate-ssl): SSL mode is off, so we won't generate a self-signed SSL key pair. 🤔 Checking for Laravel automations... ✅ Storage already linked... 🚀 Clearing Laravel cache before attempting migrations... INFO Configuration cache cleared successfully. ⚡️ Attempting connection to default database... ✅ Database connection successful RedisException ERR AUTH <password> called without any password configured for the default user. Are you sure your configuration is correct? at vendor/laravel/framework/src/Illuminate/Redis/Connectors/PhpRedisConnector.php:111 107▕ if (! empty($config['password'])) { 108▕ if (isset($config['username']) && $config['username'] !== '' && is_string($config['password'])) { 109▕ $client->auth([$config['username'], $config['password']]); 110▕ } else { ➜ 111▕ $client->auth($config['password']); 112▕ } 113▕ } 114▕ 115▕ if (isset($config['database'])) { +18 vendor frames 19 artisan:13 Illuminate\Foundation\Application::handleCommand(),

seemingly having a Redis-issue here.

Although the block looks correct, the problem is that the official redis image (redis:8-trixie) doesn't automatically read the REDIS_ARGS variable. Only the redis-stack/redis-stack-server image does that. With the normal redis image, REDIS_ARGS is set as an environment variable, but it's seemingly never passed to redis-server — the container starts with no arguments, so without requirepass. Don't know if that's an issue with one of the latter updates but this one broke my instance so I had to troubleshoot it.

Adding

command: sh -c "exec redis-server $$REDIS_ARGS"

below the REDIS_ARGS-line and deleting the contents of the redis-data-directory fixed the issue for me.

Feel free to check this out!