It is an image hosting application. It uses Redis to store metadata and filesystem to store the image itself.
This application is quite naive in terms of ranking and throttle control. Images can be liked/disliked, the number of hits is accounted but there's plenty of room to improve filtering and throttling. Throttle control is configured on 1 click per hour per image (for like/dislike but can be extended to other actions)
There's extensive use of metrics through txmetrics (https://github.com/gleicon/txmetrics) and a dashboard API/page enabled by default at <host>/api/v1/status and <host>/system_status.
You can use supervisord or another process supervisor to run imager/start.sh script or raw twistd command instead.
Install cyclone and txmetrics with pip
Clone this repository or download the tar.gz/zip file
For development and testing:
twistd -n cyclone --help
twistd -n cyclone -r imager.web.Application [--help]
For production:
twistd cyclone \
--logfile=/var/log/imager.log \
--pidfile=/var/run/imager.pid \
-r imager.web.Application
Don't forget to change the base_domain and filestore path at imager.conf. Most of the /tmp filesystems are ram based and used only for testing.
If you want to use nginx in front of it, explore the configuration files at scripts/, change the xheaders option at imager.conf to true and create a vhost on nginx that take care about the body size like this:
server {
listen 80;
server_name yourservername
access_log /var/log/nginx/yourservername.access.log;
location / {
root /var/www/yourservername/;
index index.html index.htm;
proxy_pass http://127.0.0.1:8888;
proxy_redirect off;
proxy_set_header Host $host;
proxy_set_header X-Real-IP $remote_addr;
proxy_set_header X-Forwarded-For $remote_addr;
client_max_body_size 5M;
client_body_buffer_size 5M;
proxy_connect_timeout 30;
proxy_send_timeout 30;
proxy_read_timeout 30;
proxy_buffer_size 4k;
proxy_buffers 4 32k;
proxy_busy_buffers_size 64k;
proxy_temp_file_write_size 64k;
}
}
The current cookie secret key in imager.conf was generated during the
creation of this package. However, if you need a new one, you may run the
scripts/cookie_secret.py script to generate a random key.
Add opencv image classification Generate thumbnails Group by tags