Skip to main content

Python Error Monitoring and Reporting

Project description

Flawless is a simple Python framework developed at shopkick for detecting bugs in a production environment. Flawless traps exceptions and then sends an email to the developer responsible for causing the exception. Flawless integrates with git and uses git-blame to determine which developer to email.

Project website: http://shopkick.github.com/flawless/

Why You Should Use Flawless

  • Only sends 1 email per line of code. Even if a particular line of code causes thousands of exceptions, only one email will be sent.

  • Only emails 1 developer. Flawless uses git-blame to figure out which developer is responsible for a particular exception, and will only email that developer.

  • Flawless logs the values of every variable in the stack frame at the time the exception occurred. This makes debugging ten times faster.

  • Don’t report exceptions in old code. If you set report_only_after_minimum_date, then Flawless will only report exceptions caused by code modified after report_only_after_minimum_date.

  • Don’t alert on library code. You can mark certain files/functions as library code, and when an exception originates in those files/functions, the caller will be blamed for the error instead of the library code.

4 Step Setup Guide

1. Install Flawless. After this step you should have an executable named flawless in your path.

$> python setup.py install

2. Setup the Flawless server. Go to the server which you want to host the Flawless backend. Then use the following command to start a short questionnaire to setup the server.

$> flawless configure

3. Start the Flawless server

$> flawless start -conf path/to/flawless.cfg

4. Integrate the Flawless client into your code. If you are running a WSGI application such as django or pylons/pyramid, simply add the flawless.client.middleware to your application. Otherwise you can wrap particular functions or entire classes by using flawless.client.decorators. View the examples directory for some actual code examples.

Tip: Edit flawless.client.default.py and set the default host:port for your Flawless server

How it Works

The Flawless client wraps your code with a try/except block. When an exception is caught it then sends the entire traceback to the Flawless server. The Flawless server then aggregates exception reports from clients and figures out which line of code caused the exception. Once the line that caused the exception is identified, Flawless runs “git blame” to determine the email address of the developer that last touched that line of code. Flawless then sends the developer an email with the traceback.

Exceptions can be whitelisted if they are expected. To whitelist an exception you must specify the filename, function name, and the text from the line of code being whitelisted in the appropriate config file. Alternatively, exception emails include a link to automatically add an exception to the whitelist. It is possible to whitelist all exceptions from a particular function by leaving the line of code text blank. Likewise, an entire file can be whitelisted by leaving the line of code and function blank.

Server User Interface

/get_weekly_error_report - Shows all errors that happened this week. ses a leaderboard style format to show which developer is responsible for causing the most errors this week.

Parameters:

timestamp - (optional) Specify which week you want to view. Default is the current week. include_known_errors - (optional) Include errors from config/known_errors. Default is False.

include_modified_before_min_date - (optional) Include errors originating in code modified before flawless.cfg option “report_only_after_minimum_date”. Default is False.

/check_health - Check if the server is up and running. Also displays server’s configuration parameters

/add_known_error - Webpage in which you can whitelist errors

/view_traceback - View the most recent traceback for a particular error

Parameters:

filename - (required) Specify the filename in which the error occurred function_name - (required) Specify the name of the function in which the error occurred

line_number - (required) Specify the line number on which the error occurred text - (required) Specify the full text that appears on line_number

timestamp - (optional) Specify which week you want to view. Default is the current week.

Configuration Files Reference

config/building_blocks: This is a list of library code functions that can raise an exception. Adding an entry here causes the blame to be transferred to the caller of the library rather than blaming the author of the library. See file for example.

Fields:

filename - The path to the file being whitelisted (not including the site-packages directory) function_name - The name of the function being whitelisted. This value can be set to None to act as a wildcard.

code_fragment - The actual text from the line of code being whitelisted. This value can be set to None to act as a wildcard.

config/known_errors: This is a list of known errors that happen. Reporting can be customized to completely ignore the error, to only alert after a minimum number of occurrences; or to alert every N occurences. See file for example.

Fields:

filename - The path to the file being whitelisted (not including the site-packages directory) function_name - The name of the function being whitelisted. This value can be set to None to act as a wildcard.

code_fragment - The actual text from the line of code being whitelisted. This value can be set to None to act as a wildcard.

min_alert_threshold - (optional) The minimum number of occurrences before Flawless will report this error.

max_alert_threshold - (optional) The maximum number of occurrences before Flawless will stop reporting this error

alert_every_n_occurences - (optional) Flawless will report this error every N occurrences email_recipients - (optional) List of email addresses to include on error reports for this error

email_header - (optional) Extra text to place at the top of emails for this error

config/third_party_whitelist: This is a list of errors that can be generated by thirdparty libraries that should be completely ignored (ex: network connection errors). See file for example.

Fields:

filename - The path to the file being whitelisted (not including the site-packages directory) function_name - The name of the function being whitelisted. This value can be set to None to act as a wildcard.

code_fragment - The actual text from the line of code being whitelisted. This value can be set to None to act as a wildcard

config/watched_files: This file allows developers to receive all alerts for errors related to a particular file. They can either register to receive any exception containing the file in the stacktrace, or to only receive alerts when the file is blamed for the exception. See file for example.

Fields:

email - Email address of the watcher filepath - The path to the file being watched

watch_all_errors - If true, any exception that gets reported and contains this file in it’s traceback will be sent to the watcher. If false, the watcher will only receive reports for which a line in the file was actually blamed for causing the error

config/email_remapping: Remap a developer’s email address that is returned by git-blame to instead be mapped to a different email address. See file for example.

Fields:

remap - The email address that is being remapped to - The email address that should actually receive the error reports

config/flawless.cfg: Contains all the configuration settings for the Flawless server. To view a list of configuration options, run the following command.

$> flawless options

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

flawless-0.1.8.tar.gz (45.2 kB view hashes)

Uploaded Source

Built Distributions

flawless-0.1.8.macosx-10.9-intel.tar.gz (62.6 kB view hashes)

Uploaded Source

flawless-0.1.8-py2.7.egg (87.4 kB view hashes)

Uploaded Source

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page