Application Monitoring

Each component of the application can be individually monitored. There are already some preset defaults for logging for each component.

API Monitoring

Logging

The API outputs all logs to stdout and stderr. Logs are configurable. You can adjust log level by setting the environment variable LOG_LEVEL. By default there is no timestamping of logs. This can be added by setting LOG_TIME=abs or LOG_TIME=rel.

Health

There is a singular endpoint that is publicly accessible at GET /server-health. You can use this to check if the server is up and running. This check does not verify if the database is connected to the server.

Web Monitoring

Logging

The web image is run with Caddy Server. The default caddy config is as follows

# Where caddy should listen
:2015
# Turn on the Web/file server
file_server

templates {
  mime application/javascript
  between <% %>
}
# The site root
root * /opt/app-root/www
# Because we should
encode zstd gzip


try_files {path} {path}/ index.html
header /service-worker.js {
# all static assets SHOULD be cached
  Content-Type "text/javascript"
}
  # On OCP we should log to stdout so Prometheus can
# slurp up the logs for human consumption.
log {
  errors stdout
  output stdout 
  format single_field common_log
  level info
}

Caddy will automatically surface info level logs. If you require more tuned caddy logging. You can mount your own caddy file over the original one at the path /opt/app-root/etc/.

Health Check

To test if caddy is operational just check GET / for a healthy response