The following method is used to optimize the queries in the SQuaSH API and to convert the django queryset to a pandas dataframe so that it can be used directly to make the datasources for the bokeh apps.

https://www.iwoca.co.uk/blog/2016/09/02/using-pandas-django-faster/

That's the performance results for the API endpoint that return data to feed the bokeh apps in squash, the data blobs (~50k data points) and measurement metadata after filtering by ci_id, ci_dataset and metric.

Example of API URLs, using default values:

http://localhost:8000/dashboard/api/apps/?bokeh_app=pax

and filtering by ci_id and ci_dataset

http://localhost:8000/dashboard/api/apps/?bokeh_app=pax&ci_id=841&ci_dataset=hsc

Doesn't make sent anymore since the SQuaSH RETful API was reimplemented in Flask. However a similar method was implemented for some resources of the new API (DM-13236), they return arrays in JSON that can be converted do Pandas dataframe.