Data/stats capture for analytics

Most of the information exposed through the Grafana analytics platform is obtained by running NCI tools daily on a daily basis, using the Jenkins instance running on the same VM as the Grafana tools.

The Jenkins job status is viewable here

https://accessdev.nci.org.au/jenkins/blue/organizations/jenkins/CLEX%20Admin%2Fdump_stats/activity

The scripts themselves are in this repository:

The tools use a dedicated conda environment to isolate the stat dumping from changes in the main conda/analysis3 environment.:

The environment is located here on gadi:

/g/data3/hh5/public/apps/miniconda3/envs/cms/stats

As it isn’t designed to be updated frequently, updates are done by loading the environment and manually updating installed packages.

@dale.roberts This is a start of documenting the stats dumping aspect of the analytics. I’ve changed this to a wiki post to emphasise it should be edited and updated.