Skip to content

Cue/greplin-exception-catcher

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

greplin-exception-catcher

Greplin Exception Catcher (GEC)

Exception collection and aggregation built on Google App Engine.

Why:

The ability to see aggregated exception logs in (near) real time is invaluable for determining how your application is performing.

Demonstration:

Here is a demo server running the exception catcher.

Click here to trigger 10 more fake exceptions.

The demo server generates fake exceptions using this fork.

Why App Engine:

App Engine gave us a lot of what we needed for free (task queues, persistent storage, etc) plus it makes it very easy for other companies to deploy this code base.

Status:

This is a very early stage project. It works for our needs. We haven't verified it works beyond that. Issue reports and patches are very much appreciated!

For example, some obviously needed improvements include

  • Faster data store access, particularly for large data sets.

  • Email on error spikes.

  • Over time graphs of specific errors.

  • Integration with more languages / frameworks.

  • The visual design could use a lot of love.

Pre-requisites:

App Engine

Installation

git clone https://github.com/Greplin/greplin-exception-catcher.git

Installation (server):

cd greplin-exception-catcher/server

cp example-config.json config.json

At this point, update config.json with the name and secret key. Secret key should be any random string of characters.

python setup.py

dev_appserver.py .

Notes on deploying to App Engine

You'll need to change the application identifier in app.yaml to an identifier that you own.

GEC currently requires that you attach it to a single Google Apps domain for login security.

Python using built-in logging

Installation

cd greplin-exception-catcher/python/logging

python setup.py install

Usage

import logging
from greplin.gec import logHandler

gec = logHandler.GecHandler('/path/to/exception/directory', projectName, environmentName, serverName)
gec.setLevel(logging.ERROR)
logging.getLogger().addHandler(gec)

Python using Twisted

Installation

cd greplin-exception-catcher/python/twisted

python setup.py install

Usage

from greplin.gec import twistedLog
twistedLog.GecLogObserver('/path/to/exception/directory', projectName, environmentName, serverName).start()

Javascript

  1. Create an endpoint at your server to send this stuff to GEC.
  2. Modify the call to g.errorCatcher at the end of the file to pass in functions that pass exceptions to GEC and that redact URLs respectively. (Note: your URL redaction function will be passed strings that may contain URLs, not bare URLs, so keep that in mind)
  3. Wrap your JS files if you want to capture errors during their initial execution:
     
try {
 var your_js_here 
}
      catch(e) { window.g && g.handleInitialException && g.handleInitialException(e, '(script filename here)') }
    
If you use Closure Compiler, just add this flag:
    --output_wrapper="window.COMPILED = true; try { %%output%% } catch(e) { window.g && g.handleInitialException && g.handleInitialException(e, '(script filename here)') }"
  4. This exception catching script can't see exceptions that happen before it's loaded, so make sure it's loaded early in your page before most of your other scripts.

Java using log4j

Installation

cd greplin-exception-catcher/java

mvn install

Usage

In log4j.properties:

log4j.appender.gec=com.greplin.gec.GecLog4jAppender
log4j.appender.gec.project=Project name
log4j.appender.gec.outputDirectory=/path/to/exception/directory
log4j.appender.gec.environment=prod/devel/etc.
log4j.appender.gec.serverName=Server name

Cron installation

Add the following to your crontab:

* * * * * /path/to/greplin-exception-catcher/bin/upload.py http://your.server.com YOUR_SECRET_KEY /path/to/exception/directory

Design highlights:

When exceptions occur, they are written to a directory of individual JSON files. A cron job must be set up to scan this directory for new exceptions and send them to the App Engine server.

We chose this model so that exception logging will be resilient to transient server side problems.

Authors:

Greplin, Inc.