Within the CommCell GUI, under Storage Resources > Libraries > [Library_name] is a section that displays a "Distributuion By Data Written" summary. BE VERY CAREFUL WITH THIS as it is totally misleading. Here's a screen capture from my HyperScale media agent:
It shows 158TB written but the data used to generate the pie graph only adds up to 72.6TB. It turns our that blocks that were backed up by historical jobs and still in use by current backups, are not counted if the original backup job has aged out.
From CommVault support (bolding added by me):
The jobs that aged off in the past (which means they are no longer part of the job count in the storage policy) will no longer be included in the data written calculations.
Those older aged jobs would have included baseline data and newer jobs then ran and deduped against the unique blocks laid down by those older jobs. The older job would have had a larger Data Written size, say 1TB, the subsequent job much smaller let's say 200GB. When the older larger job ages off it's unique data physically remains in the disk library in the original Volume folders because it is still being referenced by newer jobs that ran.
The "Data Written Distribution By Copies" graph and table will not show you the baseline data left by aged jobs. This is where the discrepancy arises and why the "Data Written Distribution By Copies" area of the CommVault console cannot be used as an accurate measure of deduplicated data written on a library.