Die Datenbank mit Testdaten befüllen

Warum Seed-Daten?

Von Hand geschriebene Seed-Daten

INSERT INTO users (id, email, full_name) VALUES
  (1, 'alice@example.com', 'Alice Adams'),
  (2, 'bob@example.com',   'Bob Brown');

SELECT setval('users_id_seq', (SELECT MAX(id) FROM users));

Generierte Sequenzen für große Mengen

INSERT INTO posts (author_id, title, body, published_at)
SELECT
  ((random() * 9 + 1)::int),                              -- random user 1-10
  'Post number ' || g,
  'Lorem ipsum body ' || g,
  NOW() - (random() * INTERVAL '90 days')
FROM generate_series(1, 1000) AS g;

Zufällige Fremdschlüssel

INSERT INTO comments (post_id, author_id, body)
SELECT
  (SELECT id FROM posts ORDER BY random() LIMIT 1),
  (SELECT id FROM users ORDER BY random() LIMIT 1),
  'Sample comment ' || g
FROM generate_series(1, 5000) AS g;
-- Slow at scale — see next slide for a faster pattern.

Schneller: übergeordnete Datensätze vorab zwischenspeichern

WITH u AS (SELECT id FROM users),
     p AS (SELECT id FROM posts)
INSERT INTO comments (post_id, author_id, body)
SELECT
  (SELECT id FROM p OFFSET floor(random()*1000)::int LIMIT 1),
  (SELECT id FROM u OFFSET floor(random()*10)::int   LIMIT 1),
  'Sample comment ' || g
FROM generate_series(1, 5000) AS g;

faker_fdw und Tools

COPY für große Massenimporte

COPY users (email, full_name)
  FROM '/tmp/seed_users.csv' WITH (FORMAT csv, HEADER true);

Sequenzen zurücksetzen

SELECT setval('users_id_seq', (SELECT COALESCE(MAX(id), 0) FROM users));

Idempotente Seed-Daten

INSERT INTO products (sku, name, price) VALUES
  ('A-001', 'Widget', 9.99)
ON CONFLICT (sku) DO UPDATE
  SET name = EXCLUDED.name, price = EXCLUDED.price;

Realistische Verteilung

Seed-Daten in CI

Zusammenfassung

Kurzer Test