TriggerDevelopment

From Mu2eWiki
Jump to navigation Jump to search

Intro

In this page, we present our current set of tools you can use to test the Trigger executable in the offline and online environments. We also provide some guidance the user should look at if he/she wants to develop a Filter module for the Trigger Menu.

Run the Trigger executable

The best way to run Trigger jobs is to generate your own fcl file using the script GenerateTriggerFcl. Our current best example of the full Trigger menu is implemented in the fcl Trigger/fcl/runGlobalTrigger.fcl.

What to do for creating an additional trigger path

So far, the Physics trigger filters that have been developed rely on the info from the calorimeter reconstruction or the tracking one. The filters that use only the calorimeter reconstruction are in the directory CaloFilters, while those that use the info from the tracking reconstruction are in TrkFilters. For both cases, the filters default instances and the sequences definition are in the fcl/prolog_trigger.fcl file under the respective directories. The "general" filters, like the pre-scale filter and the minimum-bias filters, are housed in the Trigger directory. In case the new filter module does not belong to one of these three categories, we will need to create a new one accordingly, so it is better to discuss it in the Trigger meeting. Then next two important steps are:

  • update the corresponding prolog_trigger.fcl file to include the new sequence required to run the new filter;
  • in case a new directory was made:
    • create a file fcl/prolog_trigger.fcl that includes the filter default instance and also the sequence required to run it;
    • include in the Trigger/fcl/templates.fcl file the new prolog_trigger.fcl file.

It is important to note that in all the trigger sequences defined so far, we have been including two event pre-scale filter instances at the very beginning and at the very end. In this way we have an handle to eventually reduce the number of times the sequence is run and/or pre-scale on the group of "good" events selected by the filter.