Debugging with Mule Flow and multiple end-points

April 26 2011

2 comments
motif

When an issue arises in production it can be quite daunting to reproduce it in a test environment.  Ideally one debugs the live application.  But logs don’t tell the whole story.  And a severe issue may require the application be taken down.  How can it be stopped and debugged at the same time?  With Mule composite sources and Mule Management Console (MMC) end-point control you can eat your cake and have it, too.

The composite-source element of a Mule can contain one or more inbound end-points.  Use one or more for production and any number of others for ; then set up the debug end-points to provide detailed visibility into your mule application.    MMC can then be used to control which end-points are enabled.  With this approach you won’t need to waste time setting up a reproduction environment when problems occur nor will you need to shut down the application to diagnose problems.

The following example augments the familiar Mule “Hello” example, applying a message-properties-transformer to add a debug flag used to manipulate control logic and application output.  Normally the application is run using port 8888.  The end-point at 18888, however, applies the transformer to set the debug flag.  A choice router in the ChitChat flow routes the message to a debugging component when the flag is detected.  This component adds flow statistics to the “Hello” output, shown here:

Below is the message-properties-transformer and composite source.  Note the “debug” flag added to the message properties and the alternate endpoint at port 18888 that applies the transformer:

This is the choice router that looks for the debug flag and routes to a debug component when the flag is set:

Inbound end-points can be stopped and started using the Mule ESB Management Console.  Navigate to the “Flows” tab, select the flow having the application end-points, then click the end-points tab.  Select the desired end-point and choose stop (or start) from the pull-down menu.

You don’t have to use the same connector type for inbound end-points.  A common scenario is to use a JMS inbound-endpoint for debugging.

The Mule ESB Management Console already provides deep runtime diagnostic capability.  Leveraging Mule composite sources and MMC end-point control further enhances diagnostic flexibility and power, with the potential to significantly reduce support turn-around time.

You can find the full Mule configuration along with the rest of the example code here.


We'd love to hear your opinion on this post


2 Responses to “Debugging with Mule Flow and multiple end-points”

  1. Now what would be excellent would be to be able to change that debug flag at runtime. For example:

    2am prod issue, no problem flip that debug switch and you can see what’s going on. With some security built it, you’re good to go!

  2. Easy, just check for the optional “debug” header which can be passed as a parameter to the URL, e.g. http://localhost:8888?debug=true
    It will be mapped to the INBOUND property.