SimCCS Map Tool

Getting Started

1. Follow the instructions for installing the Airavata Django Portal

2. With the Django Portal virtual environment activated, clone this repo and install it into the portal's virtual environment. Note, the pip install command will also run the JS frontend build and will require Node.js and Yarn installed (see the Airavata Django Portal installation instructions for more details).

   git clone https://github.com/SciGaP/simccs-maptool.git
   cd simccs-maptool
   pip install -e .
   

3. Start (or restart) the Django Portal server.

4. Open http://localhost:8000/maptool in your browser.

Django portal configuration

The following settings are relevant for the SimCCS Map Tool. These can be specified in Django Portal's settings_local.py file.

• JAVA_HOME - the Java home directory. Defaults to the JAVA_HOME env variable if not set.
• MAPTOOL_SETTINGS - this is a dictionary of Map Tool specific settings:
  • CPLEX_APPLICATION_ID - The Airavata application module id of the Cplex application to launch.
  • CPLEX_HOSTNAME - The hostname of the compute resource on which to launch Cplex.
  • DATASETS_DIR - Directory of datasets and their basedata (cost network).
  • JAVA_OPTIONS - JVM command line options. Defaults to -Xmx4g. May be a list or tuple to pass multiple options.
  • MAX_CONCURRENT_JAVA_CALLS - maximum concurrent calls into Java code allowed across all HTTP requests. Default to 1.

Example of custom settings in a settings_local.py file:

JAVA_HOME = "/usr/java/default"
MAPTOOL_SETTINGS = {
    "CPLEX_APPLICATION_ID": "Cplex_a7eaf483-ab92-4441-baeb-2f302ccb2919",
    "DATASETS_DIR": "/data/simccs-datasets"
}

Creating DB migrations

django-admin makemigrations --pythonpath . --settings simccs_maptool.tests.settings simccs_maptool

Building the Vue.js frontend code

cd frontend
yarn install
yarn run build

You can also instead run yarn run serve to start a Webpack dev server with hot reloading. See https://apache-airavata-django-portal.readthedocs.io/en/latest/dev/developing_frontend/ for more details.

Pyjnius - simccs.jar notes

Installing dependencies

In your virtual environment install the following:

pip install cython
pip install pyjnius

Building the SimCCS jar

Building simccs GitHub repo code

Note: No longer need to build. Just grab the SimCCS.jar from https://github.com/simccs/SimCCS/tree/master/store

Clone https://github.com/simccs/SimCCS

Then copy store/SimCCS.jar to simccs_maptool/simccs/lib/SimCCS.jar.

MacOS notes

I ran into issues and followed the suggestions here: joeferner/node-java#90 (comment)

Edited /Library/Java/JavaVirtualMachines/jdk1.8.0_121.jdk/Contents/Info.plist and added JNI to JVMCapabilities:

...
<key>JVMCapabilities</key>
<array>
        <string>CommandLine</string>
        <string>JNI</string>
</array>
...

Testing Pyjnius

You should be able to run the following with your virtual environment activated:

import jnius_config
import os

jnius_config.set_classpath(
    os.path.join(os.getcwd(), "simccs_maptool", "simccs", "lib", "simccs-app-1.0-jar-with-dependencies.jar"),
)
from jnius import autoclass

basepath = os.path.join(os.getcwd(), "simccs_maptool", "simccs", "Datasets")
dataset = "SoutheastUS"
scenario = "scenario1"
DataStorer = autoclass("simccs.dataStore.DataStorer")
data = DataStorer(basepath, dataset, scenario)
Solver = autoclass("simccs.solver.Solver")
solver = Solver(data)
data.setSolver(solver)
candidate_graph = data.generateCandidateGraph()