The tutorials presented here cover a range of topics, but they have in common
that they are all based on software run with ATHENA, and they all utilize
ASK.
The intend is that ASK provides enough of a buffer to be able to run the
tutorials over a variety of Atlas releases.
If you find otherwise, you should report it: <WLavrijsen@lbl.gov>,
to make sure it gets fixed.
![]() | Important notice |
|---|---|
This tutorial is "physically" part of the The text in this tutorial may seem pedantic at times, but based on feedback from users, I think this is the appropriate level. If you have run Atlas software before, you will feel every so often that I'm stating the obvious and you should just skip those parts. |
Some of the tutorials, in particular Chapter 4, Simulation tutorial, may require more disk space than you have available in, say, your home directory at CERN. Possible solutions to this problem include:
Use the Distribution Kit to install a release on your local machine.
At the time of writing, PACMAN, the main tool, is still under development,
but it works quite well, so give it a go and check back regularly.
Note that you need to define SITEROOT yourself by hand, when using ASK
with a PACMAN installation.[1]
Send a request to <atlas-support@cern.ch> and ask for scratch
space mounted on CERN AFS.
Work at Brookhaven, which has local builds of the release and you get an initial user home disk space of 2GB, plus 400MB on AFS. For details, see: www.acf.bnl.gov.
Once you have successfully run one or more of the tutorials, you will have a
properly setup environment from where you can continue working, with or
without ASK, and you will have log files that you can use for reference.
Furthermore, many terms in the tutorial are linked through to a
glossary
which, besides giving short explanations, has links to web-sites and recipes
for all of the individual steps.
This approach follows the way most people tend to work: one takes a working
environment from a colleague, bugs her for more information, and then starts
going/learning through perturbation.
Table of Contents
[1] In addition, on some platforms,
you may run into error messages from glibc that
cause ATHENA to abort.
This is because the settings on CERN Scientific Linux are lax: the problems
are real, and an abort is the only right action for
glibc to take.
Needless to say, that the software in those cases can not be used for real
work and you should report these errors for the bugs that they are.
However, if all you want to do is play around with the software, you may opt
to ignore these errors by setting the MALLOC_CHECK_ environment
variable to '0'.
You'll need this, for example, in most of the 10.x.0 and 11.0.x series because
of a bug in libshift.so.