Logging Levels0%
Basic Debugging with pdb

Logging Levels

Beginner12 min read•Updated: Jul 11, 2026
Study Materials

Logging Levels in Python

Python's logging module categorizes diagnostic and operational messages using a standardized severity hierarchy. By assigning appropriate severity levels to log messages, you can control the volume of logging output without altering your source code—viewing granular debug details during local development while filtering down to warnings and errors in production.

1. The Five Standard Logging Levels

Python defines five standard levels, each associated with an integer severity weight:

Level NameNumeric ValueWhen It Should Be Used
DEBUG10Detailed diagnostic information, variable state inspection, and function parameter tracking for developers.
INFO20Routine confirmation that operations and milestones are proceeding as expected (e.g. server booted, user signed in).
WARNING30Indication that something unexpected happened or a potential issue is emerging (e.g., deprecated API usage, low disk space), but execution continues normally.
ERROR40Due to a more serious problem, the software was unable to complete a specific task or transaction.
CRITICAL50A fatal event has occurred indicating the application or server itself may crash or be unable to continue executing.
Threshold Rule: Setting a logger's threshold to level L logs all messages with numeric value $\ge$ L, while silently suppressing everything below L.

2. Practical Examples for Every Level

Python
import logging
 
logging.basicConfig(
level=logging.DEBUG,
format="%(levelname)-8s | %(message)s"
)
 
# 1. DEBUG: Granular details
logging.debug("Connecting to database socket at 192.168.1.50:5432 with timeout=3s")
 
# 2. INFO: High-level operational events
logging.info("Database connection established successfully. Pool size: 10")
 
# 3. WARNING: Potential hazards or degradations
logging.warning("Query response took 1,420ms (exceeds 500ms SLA target)")
 
# 4. ERROR: Business transaction failure
logging.error("Failed to charge credit card for Invoice #891: Card expired")
 
# 5. CRITICAL: Unrecoverable system failure
logging.critical("Database storage drive mounted as READ-ONLY! Shutting down write worker.")

3. Capturing Exception Tracebacks: logger.exception()

When catching runtime exceptions, standard logging.error() records only your text message. To automatically include the complete Python traceback without manually formatting it, use logger.exception() (or pass exc_info=True to logger.error()):

Python
import logging
 
logging.basicConfig(level=logging.INFO)
 
def compute_ratio(numerator, denominator):
try:
return numerator / denominator
except ZeroDivisionError:
# Automatically appends the complete stack traceback to the log!
logging.exception("Division operation failed unexpectedly:")
return None
 
compute_ratio(100, 0)

Generated Output:

Output
ERROR:root:Division operation failed unexpectedly:
Traceback (most recent call last):
File "app.py", line 7, in compute_ratio
return numerator / denominator
ZeroDivisionError: division by zero

4. Dynamic Level Switching via Environment / CLI

In real-world applications, you configure the log level dynamically using environment variables or command-line flags:

Python
import logging
import os
 
# Read log level from environment variable, defaulting to INFO
env_level = os.getenv("APP_LOG_LEVEL", "INFO").upper()
 
numeric_level = getattr(logging, env_level, logging.INFO)
 
logging.basicConfig(
level=numeric_level,
format="%(asctime)s [%(levelname)s] %(message)s"
)
 
logging.debug("This only appears if APP_LOG_LEVEL=DEBUG")
logging.info("Application initialized.")

5. Summary Hierarchy Diagram

Output
Level Numeric Value Production Setting
-------------------------------------------------------
CRITICAL 50 Always recorded
ERROR 40 Always recorded
WARNING 30 Standard Production Threshold (Default)
INFO 20 Staging / Operational Audits
DEBUG 10 Local Development & Troubleshooting

Multiple Choice Questions

1. What is the default logging level if no level is explicitly configured in logging.basicConfig()?

A. DEBUG B. INFO C. WARNING D. ERROR Answer: C Explanation: By default, Python's logging module initializes with a threshold of WARNING (numeric value 30).

2. Which logging level has the lowest numeric severity value?

A. CRITICAL B. INFO C. DEBUG D. NOTSET Answer: C Explanation: Among standard operational logging levels, DEBUG has the lowest numeric value (10). (NOTSET is 0, but disables level filtering rather than serving as an operational logging level).

3. What method automatically records an ERROR level message along with the full exception traceback?

A. logging.dump_trace() B. logging.exception() C. logging.traceback() D. logging.critical_stack() Answer: B Explanation: logging.exception() logs a message at ERROR level and automatically captures and appends the active exception traceback.

4. If the active logging level is set to logging.ERROR, which of the following statements will be emitted?

A. logging.debug("Check") B. logging.info("Loaded") C. logging.warning("Slow network") D. logging.critical("System failure") Answer: D Explanation: A threshold of ERROR (40) only emits messages of severity $\ge 40$, which includes ERROR (40) and CRITICAL (50).

5. Why should passwords and sensitive API keys never be logged, even at the DEBUG level?

A. Debug logs execute slowly B. Log files are frequently ingested into centralized monitoring systems where unauthorized staff could access credentials C. The logging module fails if a string contains special symbols D. Strings longer than 10 characters cause memory leaks Answer: B Explanation: Logs are often exported, indexed, and accessible to teams. Logging sensitive credentials introduces severe security and compliance vulnerabilities.

Next Lesson

Basic Debugging with pdb

Continue learning with hands-on practice, examples, and exercises in the upcoming topic.

Related Lessons

Previous LessonNext Lesson
Using print vs loggingBasic Debugging with pdb

Practice Quiz

Test your understanding of this lesson with 5 questions. Each question has one correct answer.

PreviousNext
PrevNext