I'm trying to get DataDog to display a dashboard of system information. One piece of information is the percentage of time the system is reading/writing from the disk expressed in the metrics system.disk.read_time_pct
and system.disk.write_time_pct
However, when I put this graph on my dashboard it shows some parts at well over 5000%, which clearly can't be right.
As you can see from the preview above, it is showing a disk read time of 5430%. If I constrain the Y-axis to 100 it regularly goes above 100%.
I can't find anything to explain this or how to graph it properly. So, how do I properly graph system.disk.read_time_pct
and system.disk.write_time_pct
with DataDog?
I had issues with it too. The agent is most likely sending a value for 100 incorrectly as 10000 instead, as that's the highest I've seen it go. To make it usable, ensure the graph type is 'line', click "Advanced" on the metric, and make the equation a / 100
.
I didn't find any guidance on this either in my search, but this seems to go along with what I've seen logged locally into a Windows box and watching the performance counters locally.