Upload csv file without timestamp to kibana with ML fails

in the data visualizer of the Machine learning, we used to have a feature to upload a csv.
there is blog about that https://www.elastic.co/blog/importing-csv-and-log-data-into-elasticsearch-with-file-data-visualizer

It worked very fine in version 6 when the csv does not have a timestamp.
I migrated from version 6.5 to version 7. and now it does not work anymore if the csv does not contain any timestamp.

The error message is:

File could not be read
[illegal_argument_exception] Could not find a timestamp in the sample provided

How can I overcome this ?
this feature was very useful indeed !

For a perfectly formed CSV file with the same number of columns in every row it should still work without a timestamp in version 7.

I put more advice about this in https://github.com/elastic/kibana/issues/63526#issuecomment-738138801, but the TL;DR is that the experience for CSV files that don't have the same number of columns in every row or are corrupt in some other way is much better in 7.10.

1 Like

This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.