Skip to content

Commit

Permalink
Sergio: added README with instructions on how to run gsubbeast.
Browse files Browse the repository at this point in the history
  • Loading branch information
Sergio Maffioletti committed Mar 24, 2017
1 parent 4a75fb0 commit 40bf084
Showing 1 changed file with 98 additions and 0 deletions.
98 changes: 98 additions & 0 deletions gc3apps/usz/README.gsubbeast.txt
Original file line number Diff line number Diff line change
@@ -0,0 +1,98 @@
.. Main application developers: "Denise Kühnert"

Requirements for running ``gsubbeast``
===============================================

Takes an input folder containing 1 or more .XML files, for each of the
found .XML files runs a java application (specified in a .jar file
that could also be passed as input option_), and repeats the execution
a given number of time.

The ``gsubbeast`` command keeps a record of jobs (submitted,
executed and pending) in a session file (set name with the ``-s``
option); at each invocation of the command, the status of all recorded
jobs is updated, output from finished jobs is collected, and a summary
table of all known jobs is printed.

Running ``gsubbeast``
===============================

## Step 1: Activate ScienceCloud API authentication
$ source ~/gc3pie/bin/sc-authenticate.sh

Provide UZH Webpass shortname and password at prompt as requested.

## Step 2: Execute gsubbeast in `screen`

usage: gsubbeast [OPTIONS] input_folder

positional arguments:
input_folder Path to input folder containing valid input .xml
files.
Example:
$ screen -L `which gsubbeast` -R 2 -B ~/jar/bdmm.jar -s today -o
results -C 20 -N data/

From the provided example, one could customize:
* Location of the input folder containing the .XML files ('~/data' in
the example_).
* Location of the .jar file containing the simulation application
(change the -B option)
* Number of repetitions for each valid input .XML file (change the -R
3 option).
* Location of result folder (chage the -o results optiob).

Other GC3Pie specific options that can be customized:
* session name: the value passed to the `-s` option (`today` in the
example)

Note: the `-N` option removes all previous executions. Use it **only**
when you need to start a new simulation and forget about previous
ones.

## Step 3: Detach from running `screen` session

Once ``gsubbeast`` runs (and it has been launched with the `-C`
option, it will continuously supervise the execution of all provided
input parameters until completion. In order to properly detach from
the running session and let the ``gsubbeast`` run in background, it
is necessary to detach from the running `screen` session.

* How to detach form a running `screen` session
$ Ctrl-A Ctrl-D

## Step 4: periodically check the progress of the ``gsubbeast``
execution

Either re-attach to the running `screen` session by:
$ screen -r
Note: remember to detach from the running screen session using the:
``$ Ctrl-A Ctrl-D`` command

...or check the content of the `screen` log file:
$ less screen.log

Temporally stop a running ``gsubbeast`` session
================================================

Re-attach to the running `screen` session
$ screen -r

and stop the execution by:
$ Ctrl-C

This will interrupt the execution of ``gsubbeast`` but will **not**
terminate the instances running on ScienceCloud.

Resume ``gsubbeast`` execution
==============================
Just re-run the `gsubbeast` command. It will resume from the point it
was stopped.

Note: Unless you run `gsubbeast` command with the `-N` option. In that
case it will start a new session and loose all previous executions.

Terminating the instances on ScienceCloud started by ``gsubbeast``
===================================================================

$ gsession terminate <session path>

0 comments on commit 40bf084

Please sign in to comment.