OpenGeoscience/minerva-old

Name: minerva-old

Owner: OpenGeoscience

Description: minerva

Forked from: Kitware/minerva

Created: 2015-09-09 01:24:11.0

Updated: 2015-12-16 20:20:56.0

Pushed: 2015-11-06 23:31:44.0

Homepage:

Size: 5100

Language: Python

GitHub Committers

User	Most Recent Commit	# Commits

Other Committers

User	Email	Most Recent Commit	# Commits

README

Minerva: A Girder plugin for geospatial visualization

Assumes you have a running version of Girder.

Running Minerva

Install of system dependencies Ubuntu 14.04

This includes all system dependencies necessary for running Girder. So if you have a running Girder installation, many of these will already be satisfied.

• sudo apt-get update
• sudo apt-get install curl g++ git libffi-dev make python-dev python-pip libfreetype6-dev libpng12-dev pkg-config libgdal-dev
• sudo apt-key adv –keyserver hkp://keyserver.ubuntu.com:80 –recv 7F0CEB10
• echo 'deb http://downloads-distro.mongodb.org/repo/ubuntu-upstart dist 10gen' | sudo tee /etc/apt/sources.list.d/mongodb.list
• sudo apt-get update
• sudo apt-get install mongodb-org-server
• curl -sL https://deb.nodesource.com/setup | sudo bash -
• sudo apt-get install nodejs

Install of Minerva as a Girder plugin

• clone Minerva into the Girder plugins dir

• pip install the minerva/requirements.txt

  pip install -r requirements.txt

• run npm install in the minerva directory to get Minerva's JS dependencies

  npm install

• grunt at the top level in the girder directory to build Minerva

  grunt

• enable the Minerva plugin through the Girder Admin console

• restart Girder through the Girder Admin console

This will server Minerva as your top level application. Girder will now be served at your top level path with /girder.

Example:

Pre-Minerva:

http://localhost:8080 => serves Girder

Post-Minerva:

http://localhost:8080 => serves Minerva http://localhost:8080/girder => serves Girder

Installing a specific version of GeoJs for development

This is to get around the fact that npm is for installing packages, not managing source repositories. So when npm installs geojs, it doesn't install it as a git repo with the .git dir.

Minerva currently tracks the latest released version of geojs from npm.

If you need to use Minerva with a specific reference of geojs that isn't the current release version held in npm, do the following

1. from the minerva top level dir, cd node_modules
2. remove or move geojs, e.g. mv geojs geojs_fromnpm
3. git clone https://github.com/OpenGeoscience/geojs.git
4. cd geojs
5. checkout from git whatever branch or reference you are interested in, e.g. git checkout experimental_branch

From here, you are just following the geojs build instructions.

1. git submodule init
2. git submodule update
3. npm install
4. grunt

At this point geojs/dist/built/geo.min.js should be rebuilt, and this will be included the next time minerva is built by grunt-ing at the top level of girder. Note that if geojs/dist/* exists geojs/dist/built/geo.min.js will be used rather than geojs/geo.min.js

Setup for NEX/Spark/Romanesco mean_contour_analysis

1. From the minerva top level dir, install the mean_contour_analysis in your minerva instance

d utility
ython import_analyses.py \
ername user \
ssword password \
rt 8080 \
th /path/to/minerva/analyses/NEX/

Now you should see the mean_contour_analysis in your Analysis Panel in minerva.

1. Clone romanesco into the girder/plugins directory

d /path/to/girder/plugins
it clone https://github.com/Kitware/romanesco.git

1. install Romanesco dependencies via pip

d /path/to/girder/plugins/romanesco
ip install -r requirements.txt

1. Install Spark. You can install in a local cluster mode, and you'll want a Java 7 jvm.

e are taken from the .travis.yml file and may need to be revised.  These are more like guidelines than exact commands to run.

load spark somewhere reasonable: 
t http://psg.mtu.edu/pub/apache/spark/spark-1.3.1/spark-1.3.1-bin-hadoop2.4.tgz
r it somewhere
port SPARK_HOME=$HOME/spark-1.3.1-bin-hadoop2.4
port SPARK_MASTER_IP=localhost
the spark master
PARK_HOME/sbin/start-master.sh
the spark slave (you may need to look in the master log to find the spark address to attach to)
PARK_HOME/sbin/start-slave.sh worker1 spark://localhost:7077

1. create and configure Romanesco's worker.local.cfg in the same dir as worker.dist.cfg with content

ery]
main=romanesco
er=mongodb://localhost/romanesco

anesco]
ot dir where temp files for jobs will be written
root=tmp
mma-separated list of plugins to enable
ins_enabled=spark
lon-separated list of additional plugin loading paths
in_load_path=

1. start the Romanesco worker

d /path/to/girder/plugins/romanesco
xport SPARK_HOME=$HOME/spark-1.3.1-bin-hadoop2.4
ython -m romanesco
hings are correct it will output 'Loaded plugin "spark"'

1. grunt at the top level of Girder
2. refresh your Girder web UI page
3. enable the Romanesco plugin in the Girder web UI admin page, and restart Girder
4. configure the Romanesco plugin in Girder to allow user or group or folder access, to make it easy, and not safe for production, you can add your Girder username to the users list.