What am I doing wrong with my alert levels? Triggering early

This seems pretty straightforward but I must be missing something.

I have a metric that checks the disk usage for the root partition of a server. This is the first time setting up the system so this is a bit of a test case.

I just want to know when the disk gets nearing full. I just picked a value for now and will tweak going forward.

Is this is way more complex than I thought it was or I am making an idiot mistake. Does anyone recognize the error?

Thanks, Joe

Here is the graph for the metric

I tested earlier and was confused by the early triggering and set up again. I think that is the orange and blue triggers.

But I always thought I had my trigger values set low enough that those times should have never triggered. For instance, now I have that blue trigger set for 2.0G

screen shots for my setup in next post. noobs can’t paste more than one.

Here is the alert setup

and