We'd like to implement something similar to the "baobab" tool on Ubuntu. It's a tool that displays information about the size of all the folders and files in the hard drive.
It starts with a simple donut chart representing the aggregated size of the top-level folders. Then if one is interested, one can click on a given top-level folder, which leads to a new donut chart showing in detail the sizes of each subfolder inside that top-level folder.
Create a separate dashboard with the second level pie chart and so on
Add a drilldown to the pie chart panel to go to the second level dashboard with the selected filter (and so on)
When doing this it allows you to start on the first dashboard and drill down by clicking a pie slice of the first level, going to the dashboard with the second level and so on.
It that scalable though? The goal is to display data about files in a directory structure, for example 100's of directories. Should we create 100's of dashboards?If so, can it be done programatically?
Also, the drilldown will bring me to the same dashboard, regardless of where in the pie chart I click.
Let's say I have 2 categories in the pie chart, with 60-40 split. If I click on the item with 60%, I want to go deeper into that 60%, whereas if I click on the 40% item, I want to go deeper only in that one.
Apache, Apache Lucene, Apache Hadoop, Hadoop, HDFS and the yellow elephant
logo are trademarks of the
Apache Software Foundation
in the United States and/or other countries.