Skip to main content

​
Why did my file stop importing after I renamed it?

File from SFTP only imports files whose name and extension both match the data source’s configured filters. Renaming a file can push it outside those filters, and once it no longer matches, it is skipped on the next import. Two filters are applied when the connector lists the folder:
  • Extension. The file name must end in the configured file type (.csv). A file renamed to a different extension is skipped.
  • Name pattern. If a file-name filter is set, the file name must contain at least one of the comma-separated fragments. A rename that removes the matching fragment excludes the file.
Because each refresh re-imports all matching files and replaces the cached data, a renamed file that no longer matches drops out of the dashboard on the next scheduled refresh. To fix this, either restore a name that matches the existing filter, or update the file-name filter on the data source to include the new naming. When the file-name filter is left empty, every file with the correct extension is imported regardless of its name.

​
Why are some columns from my file missing in Adriel?

File from SFTP is free format: the available fields come from the columns of the parsed files, not a fixed catalog. A few reasons a column may not appear as expected.
The connector hides several internal columns from the available-fields list, so they never appear as selectable metrics or breakdowns: id, filename, file_updated_at, created_at, updated_at, and rowCount, along with an internal raw-breakdown placeholder. A source column that happens to share one of these reserved names is not surfaced.
Numeric columns are treated as metrics; text columns act as breakdowns. A column expected as a metric that arrives with non-numeric characters (currency symbols, thousands separators, blank cells) can be classified as text instead, so it shows up as a breakdown rather than a metric. Keeping the values clean and consistent across every file keeps the field role stable.
All matching files in the folder are concatenated into one cache table, so their columns must line up. If one file adds, drops, or reorders a column, the combined result can show missing or misaligned values. Confirm that every file delivered to the folder shares the same header row.
If the file uses a delimiter other than the one configured on the data source, the parser can read the whole row as a single column and the individual fields disappear. Set the delimiter on the data source to match the file (for example, a tab or semicolon separated export).
Available fields resolve from the cache table after the first successful import, so a newly added column appears once the file that contains it has been imported on a scheduled refresh.

​
Why is my data source connected but showing no rows?

A healthy connection only means the SFTP folder could be reached and listed. It does not guarantee that any rows were imported. Common causes:
  • No file matched the filters. If nothing in the folder matches the configured file type and file-name filter, the listing succeeds but zero files are imported. Confirm at least one .csv file in the folder matches the file-name filter.
  • The folder is empty. An empty directory is treated as a valid connection; rows only appear once a matching file is delivered.
  • The latest file has not been imported yet. Widget loads read the cache, not the live server. A file added since the last import appears only after the next scheduled refresh — see the File from SFTP data reference for the schedule.
  • The result is capped. Row queries against the cache return up to 50,000 rows. A widget expecting more than that sees a truncated result.
Because every refresh re-imports all matching files and rewrites the cache, the dashboard always reflects the files present at the last scheduled run rather than an accumulating history.

​
Can I import files from more than one folder into one data source?

No. One File from SFTP data source binds to a single upload folder, and that folder is listed flat — sub-directories are not traversed. Files nested inside sub-folders are not imported. To pull from a second folder, create a second File from SFTP data source pointed at that folder. Each data source keeps its own imported columns, so the two can be combined on a dashboard as separate sources.

​
Why can’t I group my imported data by date?

File from SFTP does not offer a native date-part breakdown, so imported rows cannot be grouped by day, week, or month the way a marketing data source can. A date column in the file is used for date-range filtering rather than as a breakdown, once it is mapped in Blend Data settings. Files without a date column are treated as a single snapshot set on the dashboard. To report by period, include the grouping value as its own text column in the file and use that column as a breakdown.

File from SFTP data reference

Import model, field type mapping, refresh strategy, and limits for File from SFTP.

Connect File from SFTP

Set up the SSH key, host, and folder for a File from SFTP data source.

Amazon S3 data reference

An alternative file source that pulls exports from an S3 bucket.

File from Email data reference

An alternative file source for inbox-delivered report attachments.