Revision 351bd07e03524f081391d68f0c6b1a1ba286a3b8 authored by carlosp420 on 05 November 2017, 15:44:47 UTC, committed by carlosp420 on 05 November 2017, 15:44:47 UTC
1 parent f33adb8
README.rst
=====
VoSeq
=====
|BuildStatus| |CoverageStatus| |Python_versions| |Chat|
.. |BuildStatus| image:: https://travis-ci.org/VoSeq/VoSeq.svg
:target: https://travis-ci.org/VoSeq/VoSeq
.. |CoverageStatus| image:: https://img.shields.io/coveralls/VoSeq/VoSeq.svg
:target: https://coveralls.io/r/VoSeq/VoSeq?branch=master
.. |Chat| image:: https://badges.gitter.im/Join%20Chat.svg
:alt: Join the chat at https://gitter.im/VoSeq/VoSeq
:target: https://gitter.im/VoSeq/VoSeq?utm_source=badge&utm_medium=badge&utm_campaign=pr-badge&utm_content=badge
.. |Python_versions| image:: https://img.shields.io/badge/python-3.4%203.5%203.6-blue.svg
:alt: Python versions
Contents
========
* `The new VoSeq is here`_
* `New features`_
* `Quick install of VoSeq using Vagrant (Recommended)`_
* `Installation instructions`_
* `Test database for development`_
* `Start a test server`_
* `Migrate VoSeq database`_
* `Set-up a publicly available web server`_
* `Administrate the server`_
* `Deployment of VoSeq`_
* `Upgrade VoSeq's software`_
* `Database backups`_
`Full documentation <http://voseq.github.io/VoSeq/>`__.
The new VoSeq is here
=====================
We have rebuilt VoSeq from scratch. We decided to migrate from PHP to
Python by using the framework Django. We also moved from MySQL to
PostgreSQL.
You can still download the old VoSeq v1.7.4 from
`here <https://github.com/VoSeq/VoSeq/releases/tag/v1.7.4>`__. But
be aware that we will not be doing major maintenance of that code.
More details about the migration can be found in our `discussion
list <https://groups.google.com/forum/#!topic/voseq-discussion-list/wQ-E0Xcimgw>`__.
VoSeq 2.0.0 is the future!
Here is a test installation of the new VoSeq (v2.0)
http://try.voseq.com/
New Features
============
Query suggestions for simple taxon searches:
.. image:: https://raw.githubusercontent.com/VoSeq/VoSeq/master/imgs/simple_search_suggestion.png
Quick install of VoSeq using Vagrant (Recommended)
==================================================
Vagrant allows setting up virtual machines that automatically installs all
dependencies and sets up configuration from a *recipe* contained in the Vagrant
file.
You need both `Vagrant <http://www.vagrantup.com/downloads.html>`__ and
`VirtualBox <https://www.virtualbox.org/wiki/Downloads>`__ installed in your
computer or server.
Assuming that you have installed [GIT](https://git-scm.com/downloads), you need to get a copy of VoSeq into your hard-disk. From a console or terminal type the following command:
.. code:: shell
git clone https://github.com/VoSeq/VoSeq.git
Just go to the VoSeq's directory:
.. code:: shell
cd VoSeq
And execute the following command:
.. code:: shell
vagrant up
If the installation of packages gets interrupted you can relaunch the process
with the following command:
.. code:: shell
vagrant reload --provision
Once the process has finished, you will have a new Ubuntu virtual machine with
VoSeq installed. To enter this virtual machine:
.. code:: shell
vagrant ssh
Then you just need to run the following commands to set up your database:
.. code:: shell
cd /vagrant
workon voseq
make migrations
Additionally, you can import your old VoSeq database from a MySQL dump (see
`Migrate VoSeq database`_). If you don't import anything your VoSeq
installation will be usable, but empty. In such a case, you might want to
import test data:
.. code:: shell
make test_import
Set up an administrator account by using the command ``make admin``
(see `Administrate the server`_).
It is necessary to index your imported data:
.. code:: shell
make index
Since this installation of VoSeq will be running as a deployed application from
inside the virtual machine you need to collect the static files in the correct
locations:
.. code:: shell
make collectstatic
Then restart the web server:
.. code:: shell
sudo supervisorctl restart voseq
sudo service nginx restart
In your host system, open your brower and load this URL:
http://33.33.33.10 to see your fresh installation of VoSeq.
Installation instructions
=========================
These instructions assume that your libraries are up to date and that you have Python, pip, Java 7+ and
virtual environments installed. Python3 is recommended.
**Step 1: get VoSeq.**
Clone or `download <https://github.com/VoSeq/VoSeq/releases>`__ VoSeq to your preferred directory.
We recommend cloning VoSeq as it will be easier to do software upgrades with on single command:
* To clone VoSeq:
.. code:: shell
git clone https://github.com/VoSeq/VoSeq.git
* To upgrade VoSeq to newer versions:
.. code:: shell
cd /path/to/VoSeq
git pull origin master
**Step 2: create a virtual environment and install dependencies.**
To ensure that all the dependencies will work without conflict, it is best to install them within a virtual environment.
.. code:: shell
mkvirtualenv -p /usr/bin/python3 voseq_environment
cd /path/to/VoSeq
workon voseq_environment
pip install -r requirements/testing.txt
Exit the virtual environment for now to continue from the shell:
.. code:: shell
deactivate
**Step 3: download and install elasticsearch.**
Elasticsearch is needed to enable the advanced search tool and auto suggestions. By default, the
elasticsearch is enabled. You can disable it in the `config.json` file by using the key:value
`"ELASTICSEARCH": "false"`. In such case, only the general search will work.
Use elasticsearch versions 1.7.3 or below. The newer versions 2.0+ currently
do not work with VoSeq.
For elasticsearch, java needs to be installed. Mac users can download and install ``elasticsearch`` from here:
http://www.elasticsearch.org/overview/elkdownloads/. In Linux, you can do:
.. code:: shell
wget https://download.elastic.co/elasticsearch/elasticsearch/elasticsearch-1.5.2.deb
sudo dpkg -i elasticsearch-1.5.2.deb
The bin directory of elasticsearch should be added automatically to your PATH. If not, add the following
line to your ``.profile`` (Linux) or ``.bash_profile`` (macOSX) file:
.. code:: shell
export PATH="$PATH:/path/to/elasticsearch/bin/"
**Step 4: download, install and configure PostgreSQL.**
For macOSX users we recommend to do it by downloading the Postgres.app from http://postgresapp.com.
Linux users can use apt-get:
.. code:: shell
sudo apt-get install postgresql postgresql-contrib postgresql-server-dev-9.3
Create new role by typing:
.. code:: shell
createuser --interactive
Enter the psql shell, create a password for this user and create a database for VoSeq:
.. code:: shell
psql
postgres=# ALTER ROLE postgres WITH PASSWORD 'hu8jmn3';
postgres=# create database voseq;
In macOSX if you are using the Postgres.app, it my be enough to run:
.. code:: shell
psql
user.name=# CREATE DATABASE voseq;
To exit the psql shell:
.. code:: shell
\q
Next, create a ``config.json`` file to keep the database variables:
.. code:: shell
cd /path/to/Voseq
touch config.json
and write in the following content:
.. code:: javascript
{
"SECRET_KEY": "create_a_secret_key",
"DB_USER": "postgres",
"DB_PASS": "hu8jmn3",
"DB_NAME": "voseq",
"DB_PORT": "5432",
"DB_HOST": "localhost",
"GOOGLE_MAPS_API_KEY": "get_a_google_map_api_key",
"PHOTOS_REPOSITORY": "local",
"ELASTICSEARCH": "true"
}
If you want to host your photos in Flickr you need to change the last parameter
of your ``config.json`` file to ``"PHOTOS_REPOSITORY": "flickr"``.
If you followed the above instructions to the letter, the DB_USER will be "postgres" and the DB_PASS
will be "hu8jmn3". It is of recommended to come up with your own password.
Instructions to obtain a personal google map browser API key can be found
`here <https://developers.google.com/maps/documentation/javascript/tutorial>`__.
You need to create a **Google Maps JavaScript API** for yourself.
After following these four steps everything should be installed and ready to run. You can now choose
to either continue with adding real data migrated from VoSeq 1.x and setting up a publicly available
web server, or to first add some test data and test the set-up with a lightweight local server
included in the VoSeq package.
Test database for development
=============================
You can use test data to populate your PostgreSQL database, useful for
development.
First, enter the virtual environment:
.. code:: shell
workon voseq_environment
Then, create tables for the database:
.. code:: shell
cd /path/to/Voseq/
make migrations
And import test data for your database:
.. code:: shell
make test_import
Start a test server
===================
In Linux start elasticsearch as a service, then enter the virtual environment and then start the server:
.. code:: shell
sudo service elasticsearch start
workon voseq_environment
cd /path/to/Voseq
make serve
In macOSX if you do not have the ``service`` command, run
``elasticsearch`` in the background and then start the server (\*):
.. code:: shell
elasticsearch -d
cd /path/to/Voseq
make serve
\* *Note that if you did not check to Start Postgres automatically after
login, you first have to go to Applications and start it manually from
there by clicking on the Postgres.app. Do this before running the
server.*
You now have a local webserver running. You can access it by opening this URL in your web browser:
``http://127.0.0.1:8000/`` and try all the buttons to see if they all work! Also notice the debug bar
on the right of the screen where you can check if all the configurations are correct.
Migrate VoSeq database
======================
If you have an existing Voseq 1.x database and want to migrate, you need to dump your MySQL database
into a XML file:
.. code:: shell
cd /path/to/Voseq/
mysqldump --xml voseq_database > dump.xml
Then use our script to migrate all your VoSeq data into a PostGreSQL
database.
.. code:: shell
make migrations
python voseq/manage.py migrate_db --dumpfile=dump.xml --settings=voseq.settings.local
If you have used a prefix for your tables in the old VoSeq, you can optionally input this as an
argument for the import script:
.. code:: shell
python voseq/manage.py migrate_db --dumpfile=dump.xml --prefix=voseq_ --settings=voseq.settings.local
It might issue a warning message:
::
WARNING:: Could not parse dateCreation properly.
WARNING:: Using empty as date for `time_edited` for code Your_Vocher_Code
It means that the creation time for your voucher was probably empty or
similar to ``0000-00-00``. In that case the date of creation for your
voucher will be empty. This will not cause any trouble when running
VoSeq. You can safely ignore this message.
Create an index for all the data in your database:
.. code:: shell
make index
If you kept your **voucher images** in your local computer or server then
your need to copy them to the correct location in the VoSeq folders:
.. code:: shell
cp old_voseq/pictures/* VoSeq/voseq/public_interface/static/.
Now copy the thumbnails of those images:
.. code:: shell
cp old_voseq/pictures/thumbnails/* VoSeq/voseq/public_interface/static/.
If you have your photos in Flickr, then don't worry you don't need to copy any
image file.
Set-up a publicly available web server
======================================
To make VoSeq available to multiple users, you will have to set-up a publicly available web server.
There are several options to do this, for example using nginx and gunicorn (best performance) or
Apache and WSGI (more suitable for hosting multiple websites).
Instructions for how to do this will follow later, but the DigitalOcean tutorials may be of use for now:
`Apache and WSGI <https://www.digitalocean.com/community/tutorials/how-to-run-django-with-mod_wsgi-and-apache-with-a-virtualenv-python-environment-on-a-debian-vps>`__
`Nginx and Gunicorn <https://www.digitalocean.com/community/tutorials/how-to-install-and-configure-django-with-postgres-nginx-and-gunicorn>`__
Administrate the server
=======================
Optionally if you want to add items/vouchers to your database
interactively, you need to create an administration account. Run the
following command and provide the requested information:
.. code:: shell
make admin
Create a database index for the simple and advanced search functions. This will speed
up the data retrieval. You need to run it once as soon as you deploy VoSeq to the server:
.. code:: shell
make index
Some features of VoSeq need to be run periodically
--------------------------------------------------
You can setup cronjobs to execute some commands once a day or every 2 hours depending on your needs.
If you remove or add data to VoSeq quite rarely (once a day) you might want to
sync the database index with your real data. In this way, new vouchers or sequences will
be found by VoSeq's search tools.
To update your database index every 24 hours (at 3:00 am) set the following cronjob:
.. code:: shell
crontab -e
Add the following line, save and exit:
.. code:: shell
0 3 * * * /path/to/python /path/to/voseq/manage.py update_index --age=24 --remove --settings=voseq.settings.production
If you add and delete data several times a day then you might want to update
your database index more often. Let us try every three hours:
.. code:: shell
0 */3 * * * /path/to/python /path/to/voseq/manage.py update_index --age=3 --remove --settings=voseq.settings.production
If you installed VoSeq using Vagrant, then your cronjob command with the correct paths should be this:
.. code:: shell
0 */3 * * * /home/vagrant/.virtualenvs/voseq/bin/python /vagrant/voseq/manage.py update_index --settings=voseq.settings.production
Update some voucher and gene statistics for your installation of VoSeq:
.. code:: shell
make stats
Deployment of VoSeq
===================
VoSeq comes with a very simple server software (from Django) that you can use
for development and testing purposes. This is the server that starts up when
you use the command ``make serve``.
However, the Django developers warn that you will need to do some extra configuration
if you want VoSeq to start serving data to the users of your lab from your institution
server or commercial servers:
* To serve statics files such as stylesheet and javascript files, you
need to choose a folder in your sever to be the root folder for such files.
Open the file ``VoSeq_repo/voseq/voseq/settings/production.py`` and change this
line so that it points to your server's folder:
.. code:: python
STATIC_ROOT = "/var/www/VoSeq/static/"
* Do something similar for being able to serve voucher images from your local
server:
.. code:: python
MEDIA_ROOT = "/var/www/VoSeq/media/"
You might want to leave it with the default values. It should work (# TODO test).
* If you have installed VoSeq in a commercial server and already bought an Internet
domain, you need to add it to the ``production.py`` file. Change the following
line:
.. code:: python
ALLOWED_HOSTS = [
'192.168.0.106', # Your Domain or IP address
]
If you don't have a domain like (myawesomedomain.com) then just replace the IP
address for the one of your server.
Before starting up VoSeq, you will need to gather all the static files in the
folders you just specified so they will be available for your users.
Use the following command:
.. code:: shell
python voseq/manage.py collectstatic --settings=voseq.settings.production
Then start VoSeq using the ``production`` configuration file:
.. code:: shell
python voseq/manage.py runserver --settings=voseq.settings.production
Upgrade VoSeq's software
========================
If you cloned the VoSeq software you can easily get the new changes by typing the following commands
in a computer terminal or console:
.. code:: shell
cd /path/to/VoSeq
git pull origin master
Do the updates to the database structure:
.. code:: shell
workon voseq_environment
make migrations
Rebuild the index and start the test server:
.. code:: shell
make index
make serve
Database backups
================
You might want to do periodical backups of your VoSeq database. You can follow these instrucctions
for backup data from postgreSQL databases: https://wiki.postgresql.org/wiki/Automated_Backup_on_Linux
Flickr Plugin
=============
VoSeq is able to host all the specimen photos in Flickr. If you have a free
account you can host up to 200 photos. The Pro account allows you hosting
unlimited number of photos for a yearly fee (25 USD).
You need to get `API keys from Flickr <https://www.flickr.com/services/api/keys/>`__
and place them in the ``config.json`` configuration file of VoSeq:
* Create and account in Flickr (if you don't own one already)
* Follow the instructions to get an API key and Secret key.
* After submitting you will get your Key and Secret. Write down those keys.
* Using a text editor software, edit the file ``config.json`` by copying your keys in it.
* For example [these are not real keys and will not work if you use them]::
.. code:: javascript
"FLICKR_API_KEY": "2d7f59f9aaa2d5c0a2782d7f5d9083a6",
"FLICKR_API_SECRET": "ef0def0f3d5f3f15f1"
* Save and exit.
Thus, every picture that you upload into your VoSeq installation will be uploaded into your Flickr account.
Computing file changes ...