Skip to content

Test Data

Three separate things, in increasing order of how much they create.

1. Initial data (needed by everyone)

cd backend
uv run manage.py load_initial_data

Loads eight fixtures in dependency order: countries, states, cities, skills, industries, qualifications, functional areas, languages.

This is reference data, not sample content — location and skill pickers are empty without it, and job posting will not work properly. Run it once after migrate.

uv run manage.py create_test_users

Creates one user of each role with predictable credentials, defined in backend/peeldb/fixtures/test-users.json.

Role Email Password
Superuser ashwin@micropyramid.com auto-generated, printed once
Company Admin ashwin.company@micropyramid.com 123456
Recruiter ashwin.recruiter@micropyramid.com 123456
Individual recruiter ashwin.individual@micropyramid.com 123456
Job seeker ashwin.jobseeker@micropyramid.com 123456

Each recruiter-type user gets five sample job posts, which is enough to exercise the dashboard.

The Company Admin and Recruiter share a company — that pairing is what you want for testing permissions, since it is the only way to check that a recruiter cannot edit another recruiter's postings. The Individual recruiter has no company at all.

Flag Effect
--clear Delete the existing test users and recreate them
--config PATH Use a different JSON definition file

Development only

These are hardcoded, publicly documented credentials. Never run this against an internet-reachable deployment.

3. Bulk sample data (for realistic volume)

uv run manage.py create_test_data

Generates data at a volume where pagination, search ranking and dashboard queries behave like production.

Flag Default
--companies 50
--recruiters 100
--jobseekers 500
--jobs 1000
--applications 3000
--clear off 
# Smaller, faster dataset
uv run manage.py create_test_data --companies=20 --recruiters=50 \
  --jobseekers=200 --jobs=500

# Wipe and regenerate
uv run manage.py create_test_data --clear

Bulk users are created with the password testpass123.

Generating the default 1000 jobs also populates the full-text search column for each — it is a generated column maintained by PostgreSQL on write, so there is no separate indexing step. See Search.

Order of operations

uv run manage.py migrate
uv run manage.py load_initial_data     # reference data — always
uv run manage.py createsuperuser       # or use create_test_users instead
uv run manage.py create_test_users     # optional: predictable logins
uv run manage.py create_test_data      # optional: volume

load_initial_data must run before either test-data command, because generated jobs reference real cities and skills.