Parse CSV Files#

CSV Enrichment

This page describes how to send a CSV file into Dynatrace for logging purposes. If instead you want to USE a CSV to enrich logs (ie. take a log and dynamically add CSV fields) then you're looking for the functionality provided by Dynatrace lookup tables

Imagine you have a CSV file that represents an ever growing list of forums posts: the date it was first posted, who asked the question and who answered it into Dynatrace. How do you do get this data into Dynatrace?

Post,Asked By,Answered,Date Posted
https://example.com/post/1,Anna,Bob,21/11/2025
https://example.com/post/2,Bob,Sarah,22/11/2025
https://example.com/post/3,Ian,Kris,22/11/2025

scenario12.yaml shows the base OpenTelemetry collector configuration we'll use during this exercise.

Stop Previous Collector#

If you haven't done so already, stop the previous collector process by pressing Ctrl + C.

Start Collector#

Run the following command to start the collector:

source /workspaces/$RepositoryName/.env
/workspaces/$RepositoryName/dynatrace-otel-collector --config=/workspaces/$RepositoryName/scenario12.yaml

Generate DataLog Data#

Open the example.csv file and add these lines then save the file.

Post,Asked By,Answered,Date Posted
https://example.com/post/1,Anna,Bob,21/11/2025
https://example.com/post/2,Bob,Sarah,22/11/2025
https://example.com/post/3,Ian,Kris,22/11/2025

View Data in Dynatrace#

Tip

Right click and "open image in new tab" to see large image

scenario12 dynatrace results

Open a new notebook (or add a DQL section to an existing notebook).

fetch logs
| search "https://example.com/post"
| fieldsKeep timestamp, content, `post`, `asked by`, `answered`, `date posted`

Click the Run button on the DQL tile. You should see the new data.

Explanation#

The filelog receiver of the collector is configured to watch for changes to example.csv.

When new content is detected, the filelog receiver ingests that data and parses it using the csv_parser operator.

We have informed the csv_parser that it should expect a header row with the named fields - if it doesn't find one, it will dynamically add those fields to the log records - meaning you can also send in CSV content without a header row.

Finally we've told the csv_parser that a comma will be the field delimiter.

The end result is a log record that looks like this as it flows through the collector (notice the data has been transformed to dedicated attributes):

LogRecord #0
ObservedTimestamp: 2025-11-12 05:29:26.4341502 +0000 UTC
Timestamp: 1970-01-01 00:00:00 +0000 UTC
SeverityText:
SeverityNumber: Unspecified(0)
Body: Str(https://example.com/post/1,Anna,Bob,21/11/2025)
Attributes:
     -> Asked By: Str(Anna)
     -> Answered: Str(Bob)
     -> Date Posted: Str(21/11/2025)
     -> log.file.name: Str(example.csv)
     -> log.file.path: Str(example.csv)
     -> Post: Str(https://example.com/post/1)
Trace ID:
Span ID:
Flags: 0

Tutorial Feedback