What is the Enterprise Data Warehouse¶
The Enterprise Data Warehouse (EDWH) provides selected, business-relevant data from Magicline as a read-only Amazon Redshift schema, updated in near real-time. It is designed for analytics — not for operating your gym, but for understanding it.
A galaxy schema built for analytics¶
The data is provided at an unaggregated, per-event grain across many fact, dimension, and bridge tables (a star / galaxy / fact-constellation schema). The schema covers members, contracts, check-ins, revenue, classes, point-of-sale, and more.
This makes it suitable for:
- BI tools such as PowerBI
- Time-uncritical direct queries
- Feeding a custom pipeline where data can be further enriched or prepared for fast access
Near real-time updates¶
Tables are synced incrementally (inserts, updates, and deletes) every few minutes, so the warehouse closely reflects the state of the source system.
Maintenance windows¶
Each region has a weekly maintenance window during which brief interruptions to updates or query availability may occur. Updates resume automatically once the window ends — no action is required on your side.
| Region | Day | Window (UTC) |
|---|---|---|
| EU | Sunday | 01:00 – 02:00 |
| US | Sunday | 07:00 – 08:00 |
| AP | Sunday | 15:00 – 16:00 |
Your warehouse's region is determined by where your tenant is hosted; check with your account manager if you're unsure which applies to you.
Change policy¶
We commit to a predictable change policy so your integrations stay stable.
Non-breaking changes¶
- New tables are announced in the Changelog on the same date they are added.
- New columns on existing tables are announced in the Changelog on the same date they are added.
Breaking changes¶
- Tables are marked deprecated in their documentation before they are deleted, so you can adapt beforehand.
- Columns are marked deprecated in their documentation before they are deleted.
- All deprecations are announced in the Changelog on the same date.
Anonymization & access levels¶
All data-privacy settings from the source system — such as anonymizing check-ins after a defined retention period — are also applied here. In addition, there are three access levels:
| Access level | What you receive |
|---|---|
| Owner | Owns the whole tenant in the source system. All data is provided. |
| Franchisee | Owns a subset of the tenant's studios. Only fact data of selected studios is provided; dimensional data of other studios may appear but is anonymized. |
| Franchise System | Owns the whole tenant but does not operate studios directly. All fact data is provided but anonymized. |
Your account's access level is defined during setup with your account manager.
Best practices¶
Use late-binding views¶
To ensure our ETL can properly refresh the underlying tables, always create views with the WITH NO SCHEMA BINDING option. A late-binding view creates no dependency on the underlying tables, so we can drop and recreate tables without breaking your views.
CREATE VIEW public.your_view_name AS
SELECT
column1,
column2
FROM
magicline.dim_or_fact_table_name
WITH NO SCHEMA BINDING;
Contact¶
For feedback and issues, please contact your account manager.