Logstash Date Parsing as Timestamp Using the Date Filter
Better Stack Team
Updated on November 18, 2024
In Logstash, you can use the date filter to parse dates from log messages and set them as the @timestamp field in your events. This is useful when the timestamp in your logs doesn’t match Logstash's default @timestamp format (which is the current time when the event is processed).
Basic Date Filter Configuration
- Use the datefilter: Extract the date from your log event.
- Specify the date format: Match it to your log's date pattern.
Example Configuration
filter {
  grok {
    match => { "message" => "\\[%{TIMESTAMP_ISO8601:log_timestamp}\\] %{LOGLEVEL:loglevel} %{GREEDYDATA:log_message}" }
  }
  date {
    match => ["log_timestamp", "ISO8601"]  # Use the appropriate format for your date
    target => "@timestamp"  # Sets the parsed date as @timestamp
    timezone => "UTC"       # Set timezone if needed (default is UTC)
  }
}
Explanation
- log_timestamp: Field created by the- grokfilter, containing the timestamp in your log entry.
- datefilter: Converts- log_timestampinto the- @timestampfield.
- match: Specifies the format. Use- "ISO8601"for standard formats, or specify a custom format (e.g.,- "dd/MMM/yyyy:HH:mm:ss Z").
- timezone: Adjust if your logs are in a timezone other than UTC.
Example Date Formats
- ISO8601: Standard format ("yyyy-MM-dd'T'HH:mm:ss.SSSZ")
- Custom Format: If your log uses a different date format, specify it (e.g., "MMM dd HH:mm:ss", "yyyy-MM-dd HH:mm:ss").
Complete Example for Custom Date
For a log entry like this:
Oct 25 2024 10:15:00 ERROR Some error message
You’d use:
filter {
  grok {
    match => { "message" => "%{MONTH:month} %{MONTHDAY:day} %{YEAR:year} %{TIME:time} %{LOGLEVEL:loglevel} %{GREEDYDATA:log_message}" }
  }
  mutate {
    add_field => { "log_timestamp" => "%{month} %{day} %{year} %{time}" }
  }
  date {
    match => ["log_timestamp", "MMM dd yyyy HH:mm:ss"]
    target => "@timestamp"
  }
}
This approach allows @timestamp to align with the original log’s timestamp, making time-based analysis more accurate.
