IKPdb Integration Guide

The idea behind integration is to configure the debugger to automatically opens a debugger client with all execution information each time an error occurs.

By default, IKPdb will open a debugger client each time an unmanaged exception is raised during program execution.

Integration will allows to do the same with programs or frameworks which includes their own exception manager.


The Odoo software integrates it’s own exception manager. As shown on the next picture, in case of error, Odoo rolls back the current transaction, opens a window displaying a stacktrace and resume execution when you click the “Ok” button.


Once integrated, when an error occurs, you automatically get this before the Odoo stack trace.


This saves you the time to open the file, find the line, set a breakpoint and restart. With integration you can start inspect variables and fix your bug straight.

Integration API

You can use 2 functions to integrate IKPDb in your code:

  • post_mortem()
  • set_trace()
ikpdb.post_mortem(trace_back=None, exc_info=None)[source]

Breaks on a traceback and send all execution information to the debugger client. If the interpreter is handling an exception at this traceback, exception information is sent to _line_tracer() which will transmit it to the debugging client. Caller can also pass an exc_info that will be used to extract exception information. If passed exc_info has precedence over traceback.

This method is useful for integrating with systems that manage exceptions. Using it, you can setup a developer mode where unhandled exceptions are sent to the developer.

Once user resumes execution, control is returned to caller. IKPdb is just used to “pretty” display the execution environement.

To call post_mortem() use:

import ikpdb
  • trace_back (traceback) – The traceback at which to break on.
  • exc_info (tuple) – Complete description of the raised Exception as returned by sys.exc_info.

An error message or None is everything went fine.

Return type:

str or None


Breaks on the line that invoked this function or at given frame. User can then resume execution.

To call set_trace() use:

import ikpdb ; ikpdb.set_trace()
Parameters:a_frame (frame) – The frame at which to break on.
Returns:An error message or None is everything went fine.
Return type:str or None

Example implementations

This list contains instructions or pointers to material describing how to integrate some system or frameworks with IKPdb.

Odoo integration

For Odoo 10

You must modify odoo/openerp/tools/debugger.py to add ikpdb in the SUPPORTED_DEBUGGER list like this:

# -*- coding: utf-8 -*-
# Part of Odoo. See LICENSE file for full copyright and licensing details.
import importlib
import logging
import types

_logger = logging.getLogger(__name__)
SUPPORTED_DEBUGGER = {'pdb', 'ipdb', 'wdb', 'pudb', 'ikpdb'}

For Odoo 8 and 9

You must modify odoo/openerp/tools/debugger.py by inserting the lines surrended by the IKPdb integration comment.

# -*- coding: utf-8 -*-
# Copyright: 2014 - OpenERP S.A. <http://openerp.com>
import types

def post_mortem(config, info):
    if config['debug_mode'] and isinstance(info[2], types.TracebackType):
            import pudb as pdb
        except ImportError:
                import ipdb as pdb
            except ImportError:
                ###( IKPdb integration begin )###
                    import ikpdb ; ikpdb.post_mortem(info)
                ###( IKPdb integration end )###
                import pdb

For Odoo 7

You must make the same modification in openerp/openerp/netsvc.py line 335.

Inouk Job Engine for Odoo

The “Inouk Job Engine” job queue management system for Odoo 9 is integrated with IKPdb . In the “Debug / Log” tab a flag allows to open IKPdb in port mortem mode if a non managed exception is raised by a job.