hose datasets Demo data with a pulse

A hose dataset · canonical

The Chinook database that didn't stop in 2013.

The same music store, the same tables you know. The invoices kept coming.

SQLite one file, ready to query CSV tar.gz, one file per table

5,154,750 rows · last row 2 minutes ago

hose → chinook 5,154,750 rows live
Rows as the generator writes them, straight from the API. Real speed. Quiet hours are quiet, the way a real business is.

The receipt

seed
42
scale
1000
start
2022-01-01
fingerprint
15c4384b1a73

The whole dataset grows from this one seed. Everyone who downloads it gets the same rows with the same ids, today and next month, so whatever you build on it keeps working after the data refreshes.

Invoices per day, last 90 days

2026-06-18: 8092026-06-19: 7772026-06-20: 14392026-06-21: 14772026-06-22: 7792026-06-23: 7702026-06-24: 7652026-06-25: 8162026-06-26: 7672026-06-27: 14542026-06-28: 14442026-06-29: 8482026-06-30: 7862026-07-01: 7842026-07-02: 7712026-07-03: 8142026-07-04: 14662026-07-05: 15002026-07-06: 8032026-07-07: 8072026-07-08: 8422026-07-09: 8112026-07-10: 8062026-07-11: 14522026-07-12: 15792026-07-13: 8872026-07-14: 8462026-07-15: 8542026-07-16: 8692026-07-17: 8352026-07-18: 15142026-07-19: 15302026-07-20: 9012026-07-21: 8862026-07-22: 8832026-07-23: 8432026-07-24: 8792026-07-25: 15992026-07-26: 15662026-07-27: 8702026-07-28: 8292026-07-29: 9242026-07-30: 9262026-07-31: 9372026-08-01: 17922026-08-02: 17152026-08-03: 9252026-08-04: 9412026-08-05: 9202026-08-06: 9972026-08-07: 9492026-08-08: 17142026-08-09: 18342026-08-10: 10222026-08-11: 10072026-08-12: 10272026-08-13: 9572026-08-14: 10292026-08-15: 19292026-08-16: 18652026-08-17: 10212026-08-18: 10382026-08-19: 11022026-08-20: 10972026-08-21: 11142026-08-22: 19702026-08-23: 19392026-08-24: 11462026-08-25: 11582026-08-26: 11542026-08-27: 12352026-08-28: 11352026-08-29: 20852026-08-30: 20742026-08-31: 11692026-09-01: 11812026-09-02: 11982026-09-03: 12032026-09-04: 12222026-09-05: 22552026-09-06: 22412026-09-07: 12672026-09-08: 12762026-09-09: 13362026-09-10: 12792026-09-11: 12162026-09-12: 22682026-09-13: 23772026-09-14: 13232026-09-15: 215

Schema · 11 tables

Artist

180 rows

  • ArtistIdint · key
  • Nametext

Album

972 rows

  • AlbumIdint · key
  • Titletext
  • ArtistIdint

Track

11,774 rows

  • TrackIdint · key
  • Nametext
  • AlbumIdint
  • MediaTypeIdint
  • GenreIdint
  • Composertext
  • Millisecondsint
  • Bytesint
  • UnitPricereal

Genre

6 rows

  • GenreIdint · key
  • Nametext

MediaType

5 rows

  • MediaTypeIdint · key
  • Nametext

Playlist

6 rows

  • PlaylistIdint · key
  • Nametext

PlaylistTrack

11,774 rows

  • idtext · key
  • PlaylistIdint
  • TrackIdint

Employee

4 rows

  • EmployeeIdint · key
  • LastNametext
  • FirstNametext
  • Titletext
  • ReportsToint
  • BirthDatetime
  • HireDatetime
  • Addresstext
  • Citytext
  • Statetext
  • Countrytext
  • PostalCodetext
  • Phonetext
  • Faxtext
  • Emailtext

Customer

53,828 rows

  • CustomerIdint · key
  • FirstNametext
  • LastNametext
  • Companytext
  • Addresstext
  • Citytext
  • Statetext
  • Countrytext
  • PostalCodetext
  • Phonetext
  • Faxtext
  • Emailtext
  • SupportRepIdint

Invoice

1,122,643 rows

  • InvoiceIdint · key
  • CustomerIdint
  • InvoiceDatetime
  • BillingAddresstext
  • BillingCitytext
  • BillingStatetext
  • BillingCountrytext
  • BillingPostalCodetext
  • Totalreal

InvoiceLine

3,953,558 rows

  • InvoiceLineIdint · key
  • InvoiceIdint
  • TrackIdint
  • UnitPricereal
  • Quantityint

Why this one

The Chinook everyone downloads was frozen in 2013. Every tutorial built on it teaches from a business where nothing has happened in over a decade, and every chart it produces falls off the same cliff.

This one runs on hose. The catalogue and the shape of the store match the original, and the sales continue through this morning. Retention curves have a today. Month-over-month comparisons include this month. Regenerate it with the same seed and every id and timestamp returns exactly as it was, so anything you build against it keeps reconciling.

The other datasets