##Batch-mobile

The webified version of Batchcave

The Environment

• Python 3.6
• Geckodriver 0.19.1 from https://github.com/mozilla/geckodriver/releases/download/v0.19.1/geckodriver-v0.19.1-linux64.tar.gz

• put it in PATH ~/anaconda3/bin or /opt

• Managed with conda.

conda env create -f environment.yml
source activate batchcave

• Batchcave environment was created with

conda install selenium=3.9.0```

### Environment Variables
* the .env file contains the variable 'DJANGOKEY' 
* run ```source .env``` from batchmobile root dir to add it to environment

## The structure
* The root-level directory, batchmobile, contains configuration files and odds and ends related to dev and testing
* batchcave is the folder for project-wide settings. No content lives here
* converter is the app that contains models, views, etc.
* why so many names? To avoid namespace issues when importing modules, avoid repeating names further up the directory structure

### Apps
* The admin app has been disabled
* converter is the app with most code and tests
* many settings are in batch_cave/settings.py

## Command Line

python3 import django django.setup() c = Conversion() c.name = 'testing' c.save() saved_c = Conversion.objects.first()

## Functional/Acceptance Tests
Functional tests are in converter/functional_tests. Run with
```python3 manage.py test functional_tests```

Tests are isolated with the LiveServerTestCase class.


## Unit Tests
Run with
```python3 manage.py test converter```


## Static Files
All static files are currently in converter/static
If there are additional apps in the future, it will be best to collect them all in to one static directory for serving. This has been configured in settings.py as /static in the root dir of the project. Collect all static filles into this folder with
```python3 manage.py collectstatic```

## The database
Flush development db with
```python3 manage.py flush```

## Set up Staging Site on a new server
* install conda
``` curl -O https://repo.continuum.io/miniconda/Miniconda3-latest-Linux-x86_64.sh
md5sum Miniconda....

Check against installer md5 at https://repo.continuum.io/miniconda/

sudo bash Miniconda...

Install in /usr/bin/miniconda3 append source to .bashrc do for local user as well as root sftp conda env file to server

conda env create -f conda_batch_cave_env.txt
source activate batchcave
cd /var/www/html/batch_cave_web
source .env
pip install pymarc

• Allow it through the firewall

sudo firewall-cmd --zone=public --add-port=8000/tcp

• Add host to ALLOWED_HOSTS in settings

• sftp the .env file to the server

• cd /var/www/html

• sudo git clone https://github.com/billmcmillin/batch_cave_web.git

• sudo mv ~/.env batch_cave_web

• python3 manage.py runserver 0.0.0.0:8000

Apache settings

ProxyPass / http://localhost:8000/
ProxyPassReverse / http://localhost:8000/ 

Setting up a new Production Site

• RHEL and Apache

source activate batchcave
sudo yum install mod_wsgi
conda install gunicorn

sudo git clone https://github.com/billmcmillin/batch_cave_web.git

sudo su tricerashopper

cd /var/www/html/batch_cave_web
gunicorn batch_cave.wsgi:application

Enable static files to be served by Apache, not Gunicorn

python manage.py collectstatic --noinput

in /etc/httpd/conf/httpd.conf: ProxyPassMatch ^/static ! ProxyPass / http://localhost:8000/ ProxyPassReverse / http://localhost:8000/

Then

sudo systemctl restart httpd.service

restart gunicorn

set a unique key on each server

in python:

a - random.SysRandom()
>>> dk = ''
>>> for i in range(0,50):
...     dk += a.choice('abcdefghijklmnopqrstuvwxyz01234567890')
... 
>>> dk

echo "conda activate" >> ~/.bashrc /etc/systemd/system/gunicorn.service

Enable Gunicorn to run as a service with Systemd

[Unit] Description=Gunicorn server for BatchCave

[Service] Restart=on-failure
User=tricerashopper WorkingDirectory=/var/www/html/batch_cave_web EnvironmentFile=/var/www/html/batch_cave_web/.env

ExecStart=/home/tricerashopper/.conda/envs/batchcave/bin/gunicorn batch_cave.wsgi:application

[Install] WantedBy=multi-user.target

---

sudo systemctl daemon-reload sudo systemctl enable gunicorn sudo systemctl start gunicorn

Run with sudo systemctl start gunicorn

• note: export command must be removed from env file for this work

##Updating

• On local machnine, copy new functions to converter/modelsdir/batchEdits.py
• run sed -i -f before_commit.sed converter/modelsdir/batchEdits.py. This removes command line only features and replaces them with web version equivalents.
• Commit to GH
• Log in to tricerashopper
• sudo systemctl stop gunicorn
• sudo su tricerashoper
• cd /var/www/html/batch_cave_web
• git pull
• exit
• sudo systemctl start gunicorn

##Fresh install on production

sudo su tricerashopper
cd ~
curl -O https://repo.continuum.io/miniconda/Miniconda3-latest-Linux-x86_64.sh
bash Miniconda....sh
source .bashrc