Server unable to start after v2.0.x update #1006
Description
Describe the bug
After upgradign from 1.17.4 -> 2.0.0, the server no longer starts and only a "504 Service Unavailable" page is returned. In the console logs on a page reload it mentions not being able to connect to the php8.4-fpm.sock socket. A little investigation returned that the php8.4-fpm service didn't start (and after trying to manually start it, does not start). I tried to look around for PHP logs but I couldn't find any, though I am totally unfamiliar with PHP and how it behaves.
Logs:
Waiting for database to be ready...
The database is now ready and reachable
sudo: unable to send audit message: Operation not permitted
[OK] Up-to-date! No migrations to execute.
Database is up to date, no migrations necessary.
AH00558: apache2: Could not reliably determine the server's fully qualified domain name, using 172.16.0.100. Set the 'ServerName' directive globally to suppress this message
AH00558: apache2: Could not reliably determine the server's fully qualified domain name, using 172.16.0.100. Set the 'ServerName' directive globally to suppress this message
[Sun Aug 31 02:35:14.105501 2025] [mpm_event:notice] [pid 1:tid 1] AH00489: Apache/2.4.62 (Debian) configured -- resuming normal operations
[Sun Aug 31 02:35:14.105709 2025] [core:notice] [pid 1:tid 1] AH00094: Command line: 'apache2 -D FOREGROUND'
[Sun Aug 31 02:35:38.151090 2025] [proxy:error] [pid 49:tid 53] (2)No such file or directory: AH02454: FCGI: attempt to connect to Unix domain socket /run/php/php8.4-fpm.sock (localhost:8000) failed
[Sun Aug 31 02:35:38.151153 2025] [proxy_fcgi:error] [pid 49:tid 53] [client 172.16.0.160:39082] AH01079: failed to make connection to backend: localhost
172.16.0.160 - - [31/Aug/2025:02:35:38 +0000] "GET / HTTP/1.1" 503 580 "-" "Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:140.0) Gecko/20100101 Firefox/140.0"
172.16.0.160 - - [31/Aug/2025:02:35:38 +0000] "GET /favicon.ico HTTP/1.1" 200 15333 "https://partdb.[DOMAIN].xyz/" "Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:140.0) Gecko/20100101 Firefox/140.0"
To Reproduce
Unsure if reproducible but my steps taken were:
- update image tag to v2.0.0 (later v2.0.1)
- wait for redeployment with updated image
- once pod is running, navigate to the Part-DB instance
Expected behavior
The server to load when navigating to it
Server Side
- Part-DB Version: 2.0.1
- PHP Version: 8.4
- Database Server: Postgres 16
- Using the Docker image in Kubernetes
Desktop (please complete the following information):
- OS: Windows 11
- Browser Waterfox
- Version 6.6.1
Additional context
Running in a Kubernetes cluster (config here )