Team, my understanding is that the 'used' field from an ONTAP CLI 'df -A' shows physical usage, per aggregate (please correct me if wrong). If I were to retrieve this programmatically via ONTAP REST using private/cli or perhaps the storage/aggregate endpoint, what field(s) should I use?
#"df -A" equivalent ONTAP REST endpoint?
weak kestrel
The reason for the uncertainty is because I'm unable to match the 'used' value from a 'df -A' to either of the values returned from a 'storage aggregate show -fields physical-used,usedsize'.
'storage aggregate show -fields usedsize' is closest, but not exact.
mossy merlin
Is compression and/or dedupe enabled?
Is cross-volume dedupe enabled (this one is a big deal)?
If so, and the system is actively used, the output of the commands/APIs at the byte or kibibyte value will change from microsecond to microsecond. I observed this behavior when supporting multiple monitoring (OnCommand) tools and customers comparing the output between tools that are on different monitoring cycles and in the Lab on Demand infrastructure where all space-saving features are enabled.
It's easier to observe these (as a human) with a more broad stroke measurement, gibibytes or tebibyte rounded to 1 or 2 decimal places.
(because I had to look up the last term... https://www.ibm.com/docs/en/sib?topic=overview-units-measurement-storage-data)
Now, that's not to say that's 100% what's happening here, but it's my experience.