Import your data
Do not start with your largest source. Sample data proves the connection works and nothing about whether it works on your material — which is where the surprises live: an unusual character set, a column that is empty half the time, a date written the American way.
Scope it as tightly as the source system allows. Read-only, if reading is all that is needed, and store it where your team already keeps secrets.
Include the environment. A connection called "test" outlives every test.
Schedule nothing until you have seen one result.
{
"name": "orders-staging",
"source": "https://data.example.com/exports/orders",
"auth": { "type": "bearer" },
"schedule": null
}Count, edges and ugly rows — the table below says what each one catches.
Fill this in: replace the example above with the real shape Fotoslov — переводчик по фото expects, and say which fields are required.
What to check, and what a failure looks like#
| Check | How | What a failure looks like |
|---|---|---|
| Count | Compare the number of records against the source | Off by a round number, usually a page limit |
| Edges | Open the oldest and the newest record | Timestamps shifted by a fixed number of hours: a time zone |
| Ugly rows | Find a record with accents, emoji or an empty field | Question marks, truncation, or a silent skip |
Do not trust a green tick on its own. An import that reports success and drops a tenth of the rows looks exactly like one that worked.
Then widen it#
Once the small source is right, add the rest one at a time, repeating the same three checks after each. Connecting six at once means debugging six at once.