Kitsu Logo

Welcome to the Kitsu API (Zou) documentation

The Kitsu API allows you to store and manage the data of your animation/VFX production. Through it, you can link all the tools of your pipeline and make sure they are all synchronized.

To integrate it into your tools, you can rely on the dedicated Python client named Gazu.

Who is it for?

The Kitsu API is made for Technical Directors, ITs, and Software Engineers from animation and VFX studios. With this API, they can enhance the tools they provide to the studio departments.

On top of the API, you can deploy the Kitsu frontend, which brings you the full bundle of the collaboration platform developed by CGWire.

Features

The Kitsu API can:

  • Store production data: projects, shots, assets, tasks, files, metadata, and validations.
  • Store preview files of any kind.
  • Publish an event stream of changes.
  • Provide folder and file paths for any task.
  • Export main data to CSV files.

For more details, you can check the full specification.

Quickstart (Docker Image)

To try Kitsu on your local machine, you can use Docker to run a local instance via this command:

Warning: This image is not aimed at production usage.

docker run -d -p 80:80 --name cgwire cgwire/cgwire

Then you can access the Kitsu API, through http://localhost/api and enjoy the Kitsu web UI at http://localhost.

The credentials are:

  • login: admin@example.com
  • password: mysecretpassword

Install

Hardware prerequisites

Users Cores RAM
1-10 2 4
11 - 30 2 8
31 - 80 4 15
81 - 200 8 30

That's the recommended minimum. But it depends on the activity of the production/studio.

  • The size of the files/videos sent
  • The frequency with which files/videos are sent
  • The network speed available between the workstations and the instance.

Regarding disk space, you need to allow for a factor of x2.5 x3 of all the files sent (large estimation).

It is advisable to separate:

  • The database on another VM
  • PREVIEW_FOLDER directory on a separate volume

This simplifies migration/augmentation of volumes.

Pre-requisites

The installation requires:

  • Ubuntu (version >= 20.04)
  • Python (version >= 3.6)
  • An up-and-running Postgres instance (version >= 9.2)
  • An up-and-running Redis server instance (version >= 2.0)
  • A Nginx instance

Setup

Dependencies

First, let's install third-party software:

sudo apt-get install postgresql postgresql-client postgresql-server-dev-all
sudo apt-get install redis-server
sudo apt-get install python3 python3-pip python3-venv
sudo apt-get install git
sudo apt-get install nginx
sudo apt-get install ffmpeg

NB: We recommend installing Postgres on a separate machine.

Get sources

Create zou user:

sudo useradd --home /opt/zou zou 
sudo mkdir /opt/zou
sudo mkdir /opt/zou/backups
sudo chown zou: /opt/zou/backups

Install Zou and its dependencies:

sudo python3 -m venv /opt/zou/zouenv
sudo /opt/zou/zouenv/bin/python -m pip install --upgrade pip
sudo /opt/zou/zouenv/bin/python -m pip install zou

Create a folder to store the previews:

sudo mkdir /opt/zou/previews
sudo chown -R zou:www-data /opt/zou/previews

Create a folder to store the temp files:

sudo mkdir /opt/zou/tmp
sudo chown -R zou:www-data /opt/zou/tmp

Prepare database

Create Zou database in Postgres:

sudo -u postgres psql -c 'create database zoudb;' -U postgres

Set a password for your postgres user. For that start the Postgres CLI:

sudo -u postgres psql

Then set the password (mysecretpassword if you want to do some tests).

psql (9.4.12)
Type "help" for help.

postgres=# \password postgres
Enter new password: 
Enter it again: 

Then, exit from the Postgres client console.

Alternatively, if you want to set the password to avoid interactive prompts, use:

sudo -u postgres psql -U postgres -d postgres -c "alter user postgres with password 'mysecretpassword';"

Finally, create database tables (it is required to leave the Postgres console and to activate the Zou virtual environment):

# Run it in your bash console.
DB_PASSWORD=mysecretpassword /opt/zou/zouenv/bin/zou init-db

NB: You can specify a custom username and database. See the configuration section.

Prepare the key-value store

Currently, Redis requires no extra configuration.

To remove warnings in Redis logs and improve background saving success rate, you can add this to /etc/sysctl.conf:

vm.overcommit_memory = 1

If you want to do performance tuning, have a look at this article.

Set up the indexer (optional)

Create a Meilisearch user:

sudo useradd meilisearch 

Install Meilisearch:

echo "deb [trusted=yes] https://apt.fury.io/meilisearch/ /" | sudo tee /etc/apt/sources.list.d/fury.list
sudo apt update && sudo apt install meilisearch

Create a folder for the index:

sudo mkdir /opt/meilisearch
sudo chown -R meilisearch: /opt/meilisearch

Define a master key then create the service file for Meilisearch:

Path: /etc/systemd/system/meilisearch.service

[Unit]
Description=Meilisearch search engine
After=network.target

[Service]
User=meilisearch
Group=meilisearch
WorkingDirectory=/opt/meilisearch
ExecStart=/usr/bin/meilisearch --master-key="masterkey"

[Install]
WantedBy=multi-user.target

To finish, start the Meilisearch indexer:

sudo service meilisearch start

Configure Gunicorn

Configure the main API server

First, create a configuration folder:

sudo mkdir /etc/zou

We need to run the application through gunicorn, a WSGI server that will run zou as a daemon. Let's write the gunicorn configuration:

Path: /etc/zou/gunicorn.conf

accesslog = "/opt/zou/logs/gunicorn_access.log"
errorlog = "/opt/zou/logs/gunicorn_error.log"
workers = 3
worker_class = "gevent"

Let's create the log folder:

sudo mkdir /opt/zou/logs
sudo chown zou: /opt/zou/logs

Then we daemonize the gunicorn process via Systemd. For that, we add a new file that will add a new daemon to be managed by Systemd:

Path: /etc/systemd/system/zou.service

Please note that environment variables are positioned here. DB_PASSWORD must be set with your database password. SECRET_KEY must be generated randomly (use pwgen 16 command for that).

[Unit]
Description=Gunicorn instance to serve the Zou API
After=network.target

[Service]
User=zou
Group=www-data
WorkingDirectory=/opt/zou
# Append DB_USERNAME=username DB_HOST=server when default values aren't used
# ffmpeg must be in PATH
Environment="DB_PASSWORD=mysecretpassword"
Environment="SECRET_KEY=yourrandomsecretkey"
Environment="PATH=/opt/zou/zouenv/bin:/usr/bin"
Environment="PREVIEW_FOLDER=/opt/zou/previews"
Environment="TMP_DIR=/opt/zou/tmp"
ExecStart=/opt/zou/zouenv/bin/gunicorn  -c /etc/zou/gunicorn.conf -b 127.0.0.1:5000 zou.app:app

[Install]
WantedBy=multi-user.target

Configure Events Stream API server

Let's write the gunicorn configuration:

Path: /etc/zou/gunicorn-events.conf

accesslog = "/opt/zou/logs/gunicorn_events_access.log"
errorlog = "/opt/zou/logs/gunicorn_events_error.log"
workers = 1
worker_class = "geventwebsocket.gunicorn.workers.GeventWebSocketWorker"

Then we daemonize the gunicorn process via Systemd:

Path: /etc/systemd/system/zou-events.service

[Unit]
Description=Gunicorn instance to serve the Zou Events API
After=network.target

[Service]
User=zou
Group=www-data
WorkingDirectory=/opt/zou
# Append DB_USERNAME=username DB_HOST=server when default values aren't used
Environment="PATH=/opt/zou/zouenv/bin"
Environment="SECRET_KEY=yourrandomsecretkey" # Same one than zou.service
ExecStart=/opt/zou/zouenv/bin/gunicorn -c /etc/zou/gunicorn-events.conf -b 127.0.0.1:5001 zou.event_stream:app

[Install]
WantedBy=multi-user.target

Configure Nginx

Finally, we serve the API through a Nginx server. For that, add this configuration file to Nginx to redirect the traffic to the Gunicorn servers:

Path: /etc/nginx/sites-available/zou

server {
    listen 80;
    server_name server_domain_or_IP;

    location /api {
        proxy_set_header Host $host;
        proxy_set_header X-Real-IP $remote_addr;
        proxy_pass http://localhost:5000/;
        client_max_body_size 500M;
        proxy_connect_timeout 600s;
        proxy_send_timeout 600s;
        proxy_read_timeout 600s;
        send_timeout 600s;
    }

    location /socket.io {
        proxy_http_version 1.1;
        proxy_set_header Host $host;
        proxy_set_header X-Real-IP $remote_addr;
        proxy_set_header Upgrade $http_upgrade;
        proxy_set_header Connection "Upgrade";
        proxy_pass http://localhost:5001;
    }
}

NB: We use the 80 port here to make this documentation simpler but the 443 port and https connection are highly recommended.

Finally, make sure that the default configuration is removed:

sudo rm /etc/nginx/sites-enabled/default

We enable that Nginx configuration with this command:

sudo ln -s /etc/nginx/sites-available/zou /etc/nginx/sites-enabled

Finally, we can start our daemon and restart Nginx:

sudo systemctl enable zou
sudo systemctl enable zou-events
sudo systemctl start zou
sudo systemctl start zou-events
sudo systemctl restart nginx

Update

Update package

First, you have to upgrade the zou package:

sudo /opt/zou/zouenv/bin/python -m pip install --upgrade zou

Update database schema

Then, you need to upgrade the database schema:

DB_PASSWORD=mysecretpassword /opt/zou/zouenv/bin/zou upgrade-db

Restart the Zou service

Finally, restart the Zou service:

sudo systemctl restart zou
sudo systemctl restart zou-events

That's it! Your Zou instance is now up to date.

NB: Make it sure by getting the API version number from https://myzoudomain.com/api.

Deploying Kitsu

Kitsu is a javascript UI that allows to manage Zou data from the browser.

Deploying Kitsu requires retrieving the built version. For that let's grab it from Github:

cd /opt/
sudo git clone -b build https://github.com/cgwire/kitsu
cd kitsu
git config --global --add safe.directory /opt/kitsu
sudo git config --global --add safe.directory /opt/kitsu
sudo git checkout build

Then we need to adapt the Nginx configuration to allow it to serve it properly:

server {
    listen 80;
    server_name server_domain_or_IP;

    location /api {
        proxy_set_header X-Real-IP $remote_addr;
        proxy_set_header Host $host;
        proxy_pass http://localhost:5000/;
        client_max_body_size 500M;
        proxy_connect_timeout 600s;
        proxy_send_timeout 600s;
        proxy_read_timeout 600s;
        send_timeout 600s;
    }

    location /socket.io {
        proxy_http_version 1.1;
        proxy_set_header Host $host;
        proxy_set_header X-Real-IP $remote_addr;
        proxy_set_header Upgrade $http_upgrade;
        proxy_set_header Connection "Upgrade";
        proxy_pass http://localhost:5001;
    }

    location / {
        autoindex on;
        root  /opt/kitsu/dist;
        try_files $uri $uri/ /index.html;
    }
}

Restart your Nginx server:

sudo systemctl restart nginx

You can now connect directly to your server IP through your browser and enjoy Kitsu!

Update Kitsu

To update Kitsu, update the files through Git:

cd /opt/kitsu
sudo git reset --hard
sudo git pull --rebase origin build

Admin users

To start with Zou you need to add an admin user. This user will be able to log in and create other users. For that go into the terminal and run the zou binary:

cd /opt/zou/
DB_PASSWORD=mysecretpassword /opt/zou/zouenv/bin/zou create-admin adminemail@yourstudio.com

It expects the password as the first argument. Then your user will be created with the email as login, default as password, and "Super Admin" as first name and last name.

Initialise data:

Some basic data are required by Kitsu to work properly (like project status) :

DB_PASSWORD=mysecretpassword /opt/zou/zouenv/bin/zou init-data

Configuration

To run properly, Zou requires a bunch of parameters you can give through environment variables. These variables can be set in your systemd script. All variables are listed in the configuration section.

Available actions

To know more about what is possible to do with the CGWire API, refer to the API section.

Packaging

Get the sources, and increment the version located in the zou/__init__.py file. Tag the repository with the new version and run the following commands:

pip install wheel twine
python setup.py bdist_wheel
twine upload dist/<package>.whl

NB: It requires access to Pypi CGWire repository.

About authors

Kitsu is written by CGWire, a company based in France. We help animation and VFX studios collaborate better through efficient tooling.

Visit cg-wire.com for more information.

CGWire Logo