Mu2e Offline Tutorial: Difference between revisions

Revision as of 23:57, 29 May 2019

Tutorial Session Goal

This tutorial will show you how to work within the Mu2e Offline computing environment.

Session Prerequisites and Advance Preparation

This tutorial may require attendees to have already completed other tutorials, and may require special preparations on their laptop. Prerequisites and advanced preparations should be listed at the top of the session, together with links. Examples might be:

Perform the tutorial on running Mu2e art Framework jobs Tutorial_2019_June:Framework_jobs
Read the root documentation on histogram fitting
Download and install root (v6.14 or later) on your machine
Run the root combined fit tutorial
Install the following docker containers ()

Session Introduction

Plan on spending 5 minutes in the beginning making sure everyone has fulfilled the prerequisites and has performed the advance preparation.
If necessary, present a brief (~5 minutes) overview on goals, background information, and context. Make sure to link the slides to the session page.

Exercises

make sure everyone has a working install of docker and has downloaded the necessary containers
- Maybe we can use a loaner computer for people having problems?
test docker using some standard containers (helloworld, bash, ...)
start an interactive docker session using the tutorial container, and explore the structure of Offline
Examine the directory tree and substructure
- Look at the different kinds of source files: .cc, .inc, .fcl, _module.cc
- Search for content (data products) in RecoDataProducts, MCDataProducts, etc
- Search for algorithms (modules) that (say) build tracks, find clusters, identify background ,...
setup a satellite release (or partial checkout?)
explore scons script and run a simple build
experiment with .build_opts
look at content in /cvmfs
look at setup.sh

Session Organization Guideline

Encourage your attendees to ask questions if they are unsure what to do.
You may or may not have session helpers, who have already performed the tutorial
If someone finishes a step or exercise early, have them look around for neighbors that need help.
If someone is stuck, have them raise their hand to get help.

Reference Materials

Use this place to add inks to reference materials.

@@ Line 17: / Line 17: @@
 == Exercises ==
-The majority of the session time should be spent performing exercises, which you link or embed in the session page.
+* make sure everyone has a working install of docker and has downloaded the necessary containers
-* Each exercise should have a title; ie ''Create a histogram of the reconstructed Ce momentum''
+** Maybe we can use a loaner computer for people having problems?
-* Exercises may be chained (ie the output of the first is used as input to the next), but in that case make sure the output is also part of the downloads, so people who failed at step 1 can continue with the subsequent steps.  Design this so that a failed step 1 does not overwrite the downloaded file that is the input to step 2.
+* test docker using some standard containers (helloworld, bash, ...)
-* Some steps may take considerable processing time.  In that case, have your attendees perform a short version (ie process 10 events), and provide the full output (ie histograms from 10^5 events) as part of the session download for subsequent steps.
+* start an interactive docker session using the tutorial container, and explore the structure of Offline
-* Each exercise should take between 5 and 15 minutes to complete.  Count on your attendees taking twice as much time as would an experienced person.
+* Examine the directory tree and substructure
-* Start each exercise by bringing the exercise up on the main screen and going through its intro.
+** Look at the different kinds of source files: .cc, .inc, .fcl, _module.cc
-* Perform each step of the exercise on the screen, describing any decisions or issues involved as you do it.  Check with your audience to make sure most (preferably all) have finished each step before going on to the next.  If someone seems stuck, assign a session helper to help them individually and move on.
+** Search for content (data products) in RecoDataProducts, MCDataProducts, etc
-* It's OK to provide more exercises than most people will complete during the tutorial; they can be completed as homework.
+** Search for algorithms (modules) that (say) build tracks, find clusters, identify background ,...
+* setup a satellite release (or partial checkout?)
+* explore scons script and run a simple build
+* experiment with .build_opts
+* look at content in /cvmfs
+* look at setup.sh
 == Session Organization Guideline ==
 * Encourage your attendees to ask questions if they are unsure what to do.