Dataset names#
Every customer has a code (e.g. katus). The environment comes first:
| Environment | Datasets |
|---|---|
| Production | prd_katus_bronze, prd_katus_silver, prd_katus_gold |
| Development | dev_katus_bronze, dev_katus_silver, dev_katus_gold |
Customers who existed before October 2026 had datasets without the prd_
(katus_silver). They were migrated; the old name keeps answering for
30 days for anyone reading the data from outside (Power BI, spreadsheets, access key).
Create Development#
In ETL/ELT → CI/CD, Environments section, use Create development environment (you must be a Production administrator). What happens:
- a sibling workspace is created, with the same owner, the same plans and its own datasets;
- the definitions are copied: pipelines, transformations, queries, dashboards,
models, agents, reports, notebooks, flows and the cluster recipes —
already pointing to the
dev_datasets; - not copied: data, connection credentials, enabled schedules, email deliveries and public links.
Ingestion in Development#
Development has its own ingestion: register its connections in ETL/ELT → Connections (they can point to a test database). Copied pipelines start paused.
Recommendation: keep Development pipelines paused and run them only when you need to test. Daily loads there double the processing cost.
Switch environments#
The Development | Production switcher sits at the top, next to the customer name. In Development, a yellow bar stays fixed on the screen. Each person only sees the switcher if they have access to both environments.
Cost#
Development usage comes out of the same balance as Production, marked as
[Desenvolvimento] on the statement.