Copyright 2011 SRI International

Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at

http://www.apache.org/licenses/LICENSE-2.0

Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.

• config

  • setup_node.py - simple script to configure a couch instance for an lr use.

• couchdb/apps

  • stemx - NSDL STEM Exchange Demo couch application

• couchdb/tests

  • stemx - scripts for testing STEM Exchange couch app
  • generate-json-evelope-data.py - OBSOLETE

• data-pumps

  • del-contents.py - simple script to delete contents of a couchdb minus design documents.
  • nsdl-to-lr-data-pump.py - OAI-PMH harvester; data pump script from NDSL-DC -> LR RDDDM
  • sample-pump-oaipmh-lib.py - OBSOLETE; here for reference.

• install

  • setup_lr.bash - a script to set up a development Learning Registry node on Ubuntu 10.04 and MacOS 10.6.x onwards.
  • install_pydeps.bash - a script to create a virtualenv and install the Learning Registry Pylons application in that virtualenv.

• LR - Pylons LR application that is used to interface w/ CouchDB.

• search - ??? A simple search UI to be installed in a design document. TODO: move into a couchapp.

• Edit the sources list

  sudo vim /etc/apt/source.list.d/sources.list

• Add deb restrticted and multiverse, plus add deb-src for all:

  deb http://us.archive.ubuntu.com/ubuntu lucid main universe restricted multiverse
  deb http://us.archive.ubuntu.com/ubuntu lucid-updates main universe restricted multiverse
  deb-src http://us.archive.ubuntu.com/ubuntu lucid main universe restricted multiverse
  deb-src http://us.archive.ubuntu.com/ubuntu lucid-updates main universe restricted multiverse

• Update apt sources

  sudo apt-get update

• Use apt-get to install curl

  sudo apt-get install libcurl3 curl

• Install Python easy_setup

  sudo apt-get install python-pkg-resources python-setuptools

• Build CouchDB dependencies

  sudo apt-get build-dep couchdb

• Install other dependencies

  sudo apt-get install xulrunner-dev libicu-dev libcurl4-gnutls-dev libtool

• Then create /etc/ld.so.conf.d/xulrunner.conf.

a. To check what XULRunner version you have installed:

  xulrunner -v

b. Configure xulrunner

i. Edit the xulrunner.conf

  sudo vi /etc/ld.so.conf.d/xulrunner.conf

ii. Add the following edits, replacing the x.x.x.x with your version number.

  /usr/lib/xulrunner-x.x.x.x
  /usr/lib/xulrunner-devel-x.x.x.x

c. Update ldconfig

  sudo /sbin/ldconfig

• Download CouchDB from http://couchdb.apache.org/downloads.html.

• Untar (decompress) the source file:

  tar -zxvf apache-couchdb-x.x.x.tar.gz

• Change into the expanded directory:

  cd apache-couchdb-x.x.x

• Install SpiderMonkey (see below)

• Configure the build:

  LDFLAGS="$(pkg-config mozilla-js --libs-only-L)" CFLAGS="$(pkg-config mozilla-js --cflags)" ./configure

• Build CouchDB:

  make

• Fix any errors

• Install CouchDB to default location:

  sudo make install

• Add couchdb user account

• change file ownership from root to couchdb user and adjust permissions

  sudo chown -R couchdb: /usr/local/var/{lib,log,run}/couchdb /usr/local/etc/couchdb
  sudo chmod 0770 /usr/local/var/{lib,log,run}/couchdb/
  sudo chmod 664 /usr/local/etc/couchdb/*.ini
  sudo chmod 775 /usr/local/etc/couchdb/*.d

• install init script and start couchdb

  cd /etc/init.d
  sudo ln -s /usr/local/etc/init.d/couchdb couchdb
  sudo /etc/init.d/couchdb start

• configure couchdb to start on system start

  sudo update-rc.d couchdb defaults

• Verify couchdb is running

  curl http://127.0.0.1:5984/

• To accesses futon remotely and run tests, update the bind address in local.ini:

  sudo vim /usr/local/etc/couchdb/local.ini
       [httpd]
       ; Bind to all addresses
       bind_address = 0.0.0.0

• Restart couchdb

  sudo service couchdb restart

• Get one of the source tarballs from http://ftp.mozilla.org/pub/mozilla.org/js/ (1.7.0 or 1.8.0-rc1 will do).

• Unpack the tarball. Note that once extracted the source are in the directory "js", without the expected version suffix.

• Go to the js/src directory.

  cd js/src

• Build SpiderMonkey. There is no default Makefile, use Makefile.ref. The default build is debug, use BUILD_OPT=1 for an optimized build.

  make BUILD_OPT=1 -f Makefile.ref

• Install SpiderMonkey. Instead of "install" the target to use is "export". Instead of PREFIX the target directory is specified with JS_DIST.

  sudo make BUILD_OPT=1 JS_DIST=/usr/local -f Makefile.ref export

• These steps are only necessary if you installed from source.

• Open logrotate.conf.

• Uncomment the the compress option, this will allow logrotate to compress old log files into a gzip file.

• run ln -s /usr/local/etc/logrotate.d/couch /etc/logrotate.d/couchdb as root or using sudo to create the couchdb entry for logrotate

• Install Yajl dependencies

  sudo apt-get install ruby cmake

• Download and extract and build Yajl 1.0.12 from here http://lloyd.github.com/yajl/

  sudo make
  sudo make install

• Update your LD_LIBRARY_PATH to include /usr/local/lib. Add the following line to ~learningregisty/.bashrc.d/defaults or /etc/bach.bashrc

  export LD_LIBRARY_PATH=/usr/local/lib:$LD_LIBRARY_PATH

• Install virtualenv

  sudo easy_install virtualenv

• Install required library dependencies

  sudo apt-get install python-dev python-libxml2 python-libxslt1 libxml2-dev libxslt1-dev

• Create a user for learningregistry and su to user

  su learningregistry

• Create a directory for the virtualenv

  mkdir virtualenv
  cd virtualenv

• Create virtualenv for LR

  virtualenv --distribute lr
  cd lr/bin/

• Install LR Python deps

  ./pip install pylons
  ./pip install flup
  ./pip install pyparsing
  ./pip install --upgrade couchdb
  ./pip install restkit
  ./pip install cython
  ./pip install lxml
  ./pip install iso8601plus
  ./pip install ijson

• Backup your original nginx.conf file

  sudo cp /etc/nginx/nginx.conf /etc/nginx/nginx.conf.bak

• Copy the ngnix.conf file from repository

  sudo cp nginx.conf /etc/nginx/nginx.conf

• Edit the production.ini found at LearningRegistry/LR

• Modify the level value under the logger_lr section

• For production is is set to only log error messages, for more verbose logging it can be set to info

• From your virtualenv directory start paster where /home/learningregistry/virtualenv/lr is my path to virtualenv.

  /home/learningregistry/virtualenv/lr/bin/paster serve --daemon production.ini start

These instructions make use of a provided installation script to install nearly all the software required to run a development Learning Registry node on Ubuntu 10.04 and MacOS 10.6.x onwards. A node set up using this script has not been tested in production. If you are feeling intrepid, do share your experiences on the mailing list.

The installation script downloads and installs the CouchBase Server Community Edition http://www.couchbase.com/downloads on Ubuntu since compilation is fiddly and this package works beautifully.

• Git, which is not available on 10.04 by default, but is available on newer versions. If you want to have the latest stable version installed you can use the Git stable releases PPA at https://launchpad.net/~git-core/+archive/ppa

• Xcode developer tools. Downloadable from the Apple Developer site at http://developer.apple.com. Currently Xcode 3 works better than Xcode 4.
• Homebrew package manager. The brew command should be in your path.
• Git, which you very likely already have if you have installed Homebrew. But can be installed from Homebrew.

Note: There is an issue installing lxml with Xcode 4, which is not required to run the Pylons app, but is required to run the tests. If Xcode 4 is installed, the script will not install lxml. See https://gist.github.com/963298 for instructions on manually installing lxml with Xcode 4.

• Clone this repository. If you are viewing this page on github, you should see the clone URL near the top of the page.

• Change directory to install directory in the cloned Learning Registry git repository.

  cd <path to clone of the git repository>/install

• Run the setup_lr.bash script. You may optionally supply the name of the virtualenv in which the python packages will be installed as an argument to the install script. If no argument is supplied, it will create a virtualenv named lr in your home directory or in the virtualenvwrapper workspace if virtualenvwrapper is installed and configured.

  bash ./setup_lr.bash [optional name of virtualenv]

• Activate the virtualenv the script created. It should have printed out instructions on how to do so.

• Start up couchdb. On Ubuntu, the CouchBase server is started up automatically after installation by the CouchBase package. You need only do the step below on MacOS.

  launchctl load -w /usr/local/Cellar/couchdb/1.0.2/Library/LaunchDaemons/org.apache.couchdb.plist

• Create a development configuration file. You may copy the original configuration file in the LR directory.

  cp LR/development.ini.orig LR/development.ini

• Navigate to the config directory within your clone of the Learning Registry git repository.

  cd config

• Load the initial documents for Learning Registry into CouchDB. This script will ask a few questions.

  ./setup_node.py

• Edit the development.ini file and replace the line use = egg:Flup#fcgi_thread with use = egg:Paste#http.

• Run the development web server.

  paster serve --reload development.ini

• Check to see if all is well by checking the status of the node we have just set up. You should receive a response if it's working.

  curl -i http://127.0.0.1:5000/status

These instructions are provided for use by developers who may be interested in running LR on MacOS 10.6.x for development purposes. MacOS has not been tested for deployment of a production Learning Registry node.

• Xcode developer tools. Downloadable from the Apple Developer site at http://developer.apple.com.
• Homebrew package manager. The brew command should be in your path.
• Git, which you very likely already have if you have installed Homebrew. But can be installed from Homebrew.
• pip.
• virtualenv.
• virtualenvwrapper.

Note: There is an issue installing lxml with Xcode 4, which is not required to run the Pylons app, but is required to run the tests. See https://gist.github.com/963298 for instructions on manually installing lxml with Xcode 4.

• Clone this repository. If you are viewing this page on github, you should see the clone URL near the top of the page.

  git clone <repository URL>

• Install CouchDB using Homebrew.

  brew install couchdb

• Create a virtualenv for Learning Registry.

  mkvirtualenv --no-site-packages --distribute learning-registry

• Install the required python packages into the virtualenv. The mkvirtualenv command will automatically activate the newly created virtualenv.

  pip install pylons pyparsing couchdb restkit lxml iso8601plus

• Change directory to the root of the cloned Learning Registry git repository.

  cd <path to clone of the git repository>

• Install the Learning Registry pylons app.

  cd LR
  pip install -e .

• Start up couchdb

  launchctl load -w /usr/local/Cellar/couchdb/1.0.2/Library/LaunchDaemons/org.apache.couchdb.plist

• Change directory to the root of the cloned Learning Registry git repository.

  cd <path to clone of the git repository>

• Create a development configuration file. You may copy the original configuration file in the LR directory.

  cp LR/development.ini.orig LR/development.ini

• Navigate to the config directory within your clone of the Learning Registry git repository.

  cd config

• Load the initial documents for Learning Registry into CouchDB. This script will ask a few questions.

  ./setup_node.py

• Edit the development.ini file and replace the line use = egg:Flup#fcgi_thread with use = egg:Paste#http.

• Run the development web server.

  paster serve --reload development.ini

• Check to see if all is well by checking the status of the node we have just set up. You should receive a response if it's working.

  curl -i http://127.0.0.1:5000/status