Unable to access Auto DevOps environments
Summary
A simple project using auto DevOps, Dockerfile based.
Unable to access the environment after auto-deploy job.
Steps to reproduce
- Create a project using auto DevOps, Dockerfile based, web server on port 5000.
- Enable Kubernetes integration (using a custom connection, my cluster is from OVH, v1.18.6)
- Install Ingress, CertManager and Prometheus on Kubernetes (using GitLab integration)
- Set the given nip.io as the base domain
- Start a review pipeline
Example Project
https://gitlab.com/geek-o-fil/sylius
What is the current bug behavior?
The auto-deploy job is giving us an URL (here http://21079362-review-wip-autode-vssxyy.51.210.210.184.nip.io/) which goes to the default backend of the Kubernetes (404).
What is the expected correct behavior?
The given URL should display the app.
Relevant logs and/or screenshots
Link to deploy job: https://gitlab.com/geek-o-fil/sylius/-/jobs/738663927
Logs from main container:
2020-09-15T12:39:14.805222335Z [Web Server/PHP ] Sep 15 12:39:14 |DEBUG| PHP started
2020-09-15T12:39:14.805571110Z [Web Server/PHP ] Sep 15 12:39:14 |INFO | PHP fpm is running, pid 53
2020-09-15T12:39:14.805895745Z [Web Server/PHP ] Sep 15 12:39:14 |INFO | PHP ready to handle connections
2020-09-15T12:39:15.571500824Z
2020-09-15T12:39:15.571624990Z [OK] Web server listening
2020-09-15T12:39:15.571670908Z The Web server is using PHP FPM 7.3.22
2020-09-15T12:39:15.571709759Z http://127.0.0.1:5000
2020-09-15T12:39:15.571723484Z
2020-09-15T12:39:18.811160341Z [Web Server/PHP ] Sep 15 12:39:18 |INFO | SERVER GET (302) /
2020-09-15T12:39:19.059666782Z [Web Server/PHP ] Sep 15 12:39:18 |INFO | SERVER GET (200) /en_US/
Containers state after deploy job:
Output of checks
This bug happens on GitLab.com
Results of GitLab environment info
Not relevant.
Results of GitLab application Check
Not relevant.
Possible fixes
//
Edited by Quentin “Storm1er” Decaunes