I fired up the debugger and watched the code happen tonight.
The problem occurs if the config.action_controller.
perform_caching = true in your environment. It’s not set to true in development but it is set to true in production and staging.
The problem occurs not in my controllers but in sweeping.rb (part of rails). Rails calls this code for every controller which was instantiated in the same request ( I think). Running the page it “sweeps” the following controllers in this order.
#<MatchCommissionedScheduledController:0xb679b4fc> #MatchController:0xb61694f8
The problem is that on the second call to clean up MatchController it pukes. The code snippets in sweeping.rb are…
The cleanup is in the following method of sweep.rb
def after(controller) callback(:after) if controller.perform_caching # Clean up, so that the controller can be collected after this request
self.controller = nil
end
A little later…
def callback(timing) controller_callback_method_name = “#{timing}_#{controller.controller_name.underscore}”
action_callback_method_name = "#{controller_callback_method_name}_#{controller.action_name}"
__send__(controller_callback_method_name) if respond_to?(controller_callback_method_name, true)
__send__(action_callback_method_name) if respond_to?(action_callback_method_name, true)
end
The problem is that after the first controller has been cleared controller is nil when it gets to the callback (even if it’s not nil when it hits the after method)
I am not quite sure what is the problem but could this be a bug in sweeper.rb?