Errors after rebranding from GamerHub to SaaSight:

ERROR-FIX LOG (Post-Rebrand)

1️⃣ Problem: Wrong old DB file (boardgames.db) kept getting recreated

• Symptom: Homepage and register/login worked until you deleted boardgames.db. Then app crashed.
• Fix:

  • Updated config.py → Changed:

    'sqlite:///' + os.path.join(os.path.abspath(os.path.dirname(__file__)), 'boardgames.db')
    
    

    to:

    'sqlite:///' + os.path.join(os.path.abspath(os.path.dirname(__file__)), 'saasight.db')
    
    

  • Pushed changes → pipeline now uses correct DB name.

2️⃣ Problem: CodePipeline was deploying old artifacts

• Symptom: New commits triggered build, but EC2 didn’t reflect changes.
• Fix:
  • Ensured SourceArtifact → BuildArtifact → Deploy chain is correct.
  • Added missing OutputArtifact: BuildArtifact in Build Stage.
  • Fixed CodePipeline input/output references.
  • Verified in S3 that latest zips had updated content.
  • Cleaned /home/ec2-user/ in install script to remove stale files.

3️⃣ Problem: Gunicorn systemd service failed (Exec path broken)

• Symptom: EC2 didn’t serve app. systemctl status saasight showed No such file or directory for Gunicorn path.
• Fix:
  • Recreated virtualenv in install script after copying latest code.
  • Regenerated venv/bin/gunicorn path properly.
  • Restarted saasight.service with valid Gunicorn path in ExecStart.

4️⃣ Problem: EC2 still served stale version (e.g., old base.html)

• Symptom: curl localhost:8000 returned old templates despite new commits.
• Fix:

  • Validated CodeDeploy archive via:

    find /opt/codedeploy-agent/deployment-root -name base.html
    
    

  • Ensured install_dependencies.sh pulls latest from newest deployment-archive using:

    find ... | sort | tail -1
    
    

  • Cleaned all relevant folders before copying.