Installation

edit
$ python -m pip install ecs-logging

Getting Started

edit

ecs-logging-python has formatters for the standard library logging module and the structlog package.

Standard Library logging Module

edit
import logging
import ecs_logging

# Get the Logger
logger = logging.getLogger("app")
logger.setLevel(logging.DEBUG)

# Add an ECS formatter to the Handler
handler = logging.StreamHandler()
handler.setFormatter(ecs_logging.StdlibFormatter())
logger.addHandler(handler)

# Emit a log!
logger.debug("Example message!", extra={"http.request.method": "get"})
{
    "@timestamp": "2020-03-20T18:11:37.895Z",
    "log.level": "debug",
    "message": "Example message!",
    "ecs": {
        "version": "1.6.0"
    },
    "http": {
      "request": {
        "method": "get"
      }
    },
    "log": {
        "logger": "app",
        "origin": {
            "file": {
                "line": 14,
                "name": "test.py"
            },
            "function": "func"
        },
        "original": "Example message!"
    }
}

Excluding Fields

edit

You can exclude fields from being collected by using the exclude_fields option in the StdlibFormatter constructor:

from ecs_logging import StdlibFormatter

formatter = StdlibFormatter(
    exclude_fields=[
        # You can specify individual fields to ignore:
        "log.original",
        # or you can also use prefixes to ignore
        # whole categories of fields:
        "process",
        "log.origin",
    ]
)

Limiting Stack Traces

edit

The StdlibLogger automatically gathers exc_info into ECS error.* fields. If you’d like to control the number of stack frames that are included in error.stack_trace you can use the stack_trace_limit parameter (by default all frames are collected):

from ecs_logging import StdlibFormatter

formatter = StdlibFormatter(
    # Only collects 3 stack frames
    stack_trace_limit=3,
)
formatter = StdlibFormatter(
    # Disable stack trace collection
    stack_trace_limit=0,
)

Structlog Example

edit

Note that the structlog processor should be the last processor in the list, as it handles the conversion to JSON as well as the ECS field enrichment.

import structlog
import ecs_logging

# Configure Structlog
structlog.configure(
    processors=[ecs_logging.StructlogFormatter()],
    wrapper_class=structlog.BoundLogger,
    context_class=dict,
    logger_factory=structlog.PrintLoggerFactory(),
)

# Get the Logger
logger = structlog.get_logger("app")

# Add additional context
logger = logger.bind(**{
    "http": {
        "version": "2",
        "request": {
            "method": "get",
            "bytes": 1337,
        },
    },
    "url": {
        "domain": "example.com",
        "path": "/",
        "port": 443,
        "scheme": "https",
        "registered_domain": "example.com",
        "top_level_domain": "com",
        "original": "https://example.com",
    }
})

# Emit a log!
logger.debug("Example message!")
{
  "@timestamp": "2020-03-26T13:08:11.728Z",
  "ecs": {
    "version": "1.6.0"
  },
  "http": {
    "request": {
      "bytes": 1337,
      "method": "get"
    },
    "version": "2"
  },
  "log": {
    "level": "debug"
  },
  "message": "Example message!",
  "url": {
    "domain": "example.com",
    "original": "https://example.com",
    "path": "/",
    "port": 443,
    "registered_domain": "example.com",
    "scheme": "https",
    "top_level_domain": "com"
  }
}

Elastic APM Log Correlation

edit

ecs-logging-python supports automatically collecting ECS tracing fields from the Elastic APM Python agent in order to correlate logs to spans, transactions and traces in Elastic APM.

You can also quickly turn on ECS-formatted logs in your python app by setting LOG_ECS_REFORMATTING=override in the Elastic APM Python agent.

Install Filebeat

edit

The best way to collect the logs once they are ECS-formatted is with Filebeat:

  1. Follow the Filebeat quick start
  2. Add the following configuration to your filebeat.yaml file.

For Filebeat 7.16+

filebeat.yaml.

filebeat.inputs:
- type: filestream 
  paths: /path/to/logs.json
  parsers:
    - ndjson:
      overwrite_keys: true 
      add_error_key: true 
      expand_keys: true 

processors: 
  - add_host_metadata: ~
  - add_cloud_metadata: ~
  - add_docker_metadata: ~
  - add_kubernetes_metadata: ~

Use the filestream input to read lines from active log files.

Values from the decoded JSON object overwrite the fields that Filebeat normally adds (type, source, offset, etc.) in case of conflicts.

Filebeat adds an "error.message" and "error.type: json" key in case of JSON unmarshalling errors.

Filebeat will recursively de-dot keys in the decoded JSON, and expand them into a hierarchical object structure.

Processors enhance your data. See processors to learn more.

For Filebeat < 7.16

filebeat.yaml.

filebeat.inputs:
- type: log
  paths: /path/to/logs.json
  json.keys_under_root: true
  json.overwrite_keys: true
  json.add_error_key: true
  json.expand_keys: true

processors:
- add_host_metadata: ~
- add_cloud_metadata: ~
- add_docker_metadata: ~
- add_kubernetes_metadata: ~

For more information, see the Filebeat reference.