The data-flow monitor is designed to check the SNO+ Grafana page for any issues in transfer efficiency, and report them both daily and weekly to the processing group.
As of v2.0, the tool no longer relies on scraping as I was able to finally acquire an API key. This allows the tool to run significantly faster and more reliably as it can simply make a call to the data source and retrieve the result rather than running through all of the webpages.
- virtualenv (in order to install the isolated python environment)
- Download the release directory
- Modify run_script.sh to ensure it changes into the correct install directory so that Cron can see the contents.
- Add execute permissions to run_script.sh (chmod +x run_script.sh)
- Ensure you have the grafana API token stored somewhere and modify run_script.sh to pass in the path to it when running data_processor.py
- (OPTIONAL) Schedule as a Cron job so that it runs on a schedule.
./run_script
Currently, our Grafana is posting data one hour behind (ex. if it is 14:23, then it should post 14:00 as the latest but it posts 13:00) so there is a line in the data_processor.py to account for this. This is potentially by design from whoever set up our dashboard, so this may need to be modified depending on how yours is set up and will probably become an option in the config soon.