CSV

The CSV source tails CSV-formatted log files on the collector host and parses each row into a structured log record. You supply a comma-delimited header that names the columns, and Bindplane maps each value in a row to the matching key.

Supported Telemetry Types

Platform	Metrics	Logs	Traces
Linux		✓
Windows		✓
macOS		✓

Prerequisites

• A collector installed on the host where the CSV files reside.
• Read access for the collector to the file or directory paths being tailed.
• A header that lists the column names in the same order they appear in the files.

Configuration

General

Parameter	Type	Default	Description
Header	String		A comma-delimited list of keys assigned to each of the columns. Required.
File Path(s)	Strings	[]	File or directory paths to tail for logs. Required.
Exclude File Path(s)	Strings	[]	File or directory paths to exclude from tailing.
Log Type	String	csv	A friendly name that is added to each log entry as an attribute.

Advanced

Parameter	Type	Default	Description
Encoding	Enum	utf-8	The encoding of the file being read. Valid values: nop, utf-8, utf-16le, utf-16be, ascii, big5.
Start At	Enum	end	Start reading logs from beginning or end of the file.
Parse	Boolean	true	Parses the CSV fields into structured data using the configured header. Set to false to keep the raw line without parsing.

Example Configuration

Standalone Source

apiVersion: bindplane.observiq.com/v1
kind: Source
metadata:
  id: csv
  name: csv
spec:
  type: csv
  parameters:
    - name: header
      value: 'timestamp,level,method,path,status'
    - name: file_path
      value:
        - '/var/log/myapp/*.csv'
    - name: log_type
      value: 'myapp'
    - name: start_at
      value: 'end'
    - name: encoding
      value: 'utf-8'
    - name: parse
      value: 'true'

Configuration Tips

Matching the header to your files

• List the keys in header in the exact order the columns appear in the file. The number of keys should match the number of columns, otherwise rows may parse with missing or shifted values.
• If your files do not include a header row, the header parameter still supplies the column names, so the first data row is parsed normally.

File paths and read position

• file_path accepts both individual files and directory glob patterns (for example, /var/log/myapp/*.csv), and multiple entries can be supplied.
• Use start_at: beginning to ingest existing file contents on first run. Leave it at end to read only new lines appended after the collector starts.
• Use exclude_file_path to skip files that would otherwise match a file_path glob, such as rotated or archived files.

Encoding

• Set encoding to match the file's character set. Use nop to read bytes without decoding when the file is not text.

Troubleshooting

Symptom: Logs are ingested but the fields are not split into separate attributes. Solution: Confirm parse is set to true and that header lists every column. When parsing is disabled or the header is empty, rows are kept as the raw line instead of structured data.

Symptom: Column values appear under the wrong keys. Solution: Verify the order and count of keys in header match the columns in the file. A mismatch shifts values into adjacent fields.

Symptom: No logs are collected from existing files. Solution: The source defaults to start_at: end, which reads only new lines. Set start_at to beginning to ingest content that already exists in the files, and confirm the collector has read access to the configured paths.