create and activate your virtualenv - use format live|backup|dev plus the date

for example, backup-14-05-22

virtualenv backup-14-05-22 cd backup-14-05-22 source bin/activate

clone this project into the virtualenv

git clone https://github.com/evildmp/arkestra_medic.git

some packages use the wheel installer, so make it available

pip install wheel

install all the packages it needs

pip install -r arkestra_medic/requirements.txt

don't forget to install the Arkestra and Arkestra Publications packages that need

to be installed from source

pip install -r src/arkestra/REQUIREMENTS.txt pip install -r src/arkestra-publications/REQUIREMENTS.txt

copy the database; give it the same name as the virtualenv

copy (from an existing installation) and modify the secret settings file as required

check that the database has the correct permissions for the user

check that the runserver works

python arkestra_medic/manage.py runserver 0.0.0.0:8000

gather the static media files ready to be served by the production webserver

python arkestra_medic/manage.py collectstatic -l

edit the restart_fgci file so that it refers to the virtualenv name

start the project

arkestra_medic/restart-fcgi start

it may be necessary to install flup for the previous step to work

pip install flup

edit the fcgi.conf (this is an nginx configuration) file:

change instances of /path/to/virtualenv to match the virtualenv's path

so that the upstream name matches that of the virtualenv

so that the list (i.e. port) and server_name values are correct

add a line to /opt/nginx/conf/nginx.conf to include this configuration, for example:

include /home/daniele/backup-14-05-21/arkestra_medic/fastcgi.conf;

restart nginx

sudo /etc/init.d/nginx restart

copy the site's media files (uploaded and processed images and video and other

files) from the existing installation:

arkestra_medic/arkestra_medic/media/filer_public

arkestra_medic/arkestra_medic/media/filer_public_thumbnails

arkestra_medic/arkestra_medic/media/rendered_video

edit the arkestra_medic/symplectic-scripts file to reflect the correct virtualenv name

edit /etc/crontab to run the correct scripts at the correct time

for example:

15 07 * topdog /path/to/name-of-virtualenv/arkestra_medic/symplectic-scripts

15 19 * topdog /path/to/name-of-virtualenv/arkestra_medic/symplectic-scripts

for live sites only (not training etc)

find the server's database backup script and edit its DBNAME to match the live database name

this script dumps the database so it can be backed up by legato

on live, this script is at /home/topdog/backups/arkestra-db-dump.sh; cron runs it daily