Slack bot that shares and promotes awareness of KA's testimonials
To run locally:
- First run
make secrets.py. - Then run
make deps - Then run
make serve
== Deployment
There's two separate things to deploy here:
The api endpoints that send new testimonials to slack runs as a appengine
standard application. Deploy these by running make deploy. To test that the
deploy succeeded, you can run:
curl http://khan-testimonials-turtle.appspot.com/test_new_testimonial
and verify that a new default testimonial is sent to the #testimonials
channel.
For access to the App Engine part of this application, ask #it for access to khan-testimonials-turtle GCP project.
The slack client that listens for reactions to slack messages and updates vote
totals (listener.py) runs on toby, using the configuration found
here.
To update that code, follow these steps:
# SSH into toby:
gcloud --project=khan-internal-services compute ssh \
ubuntu@toby-internal-webserver
# cd to the project directory
cd /home/ubuntu/slack-testimonials-bot
# pull the latest code
git fetch && git checkout origin/master
# restart the service to pull in that latest code
sudo systemctl restart slack-testimonials-bot
To test that this deploy succeeded, upvote a recently high quality testimonial.
This should eventually reshare that testimonial to the #khan-academy channel.
== Fixing when it breaks
If something is weird with the App Engine deployment, try going in to cloud storage and wiping the staging bucket in there.