Problem/Justification
Currently to monitor the temp on an HBA, I need to do this via a shell. either running a command, or viewing a log that gets populated.
It got me thinking, how hard would it be to add functionality to the existing report section to allow for custom charts?
Impact
I can’t see any negative impacts as it would not impact existing functionality. It would simply add a custom option (and perhaps customisable sub sections) to the existing drop down lists.
There are options for going a bit further and having a group of reports called custom and then having heading + sub headings so that it functions like the disk reports where you can filter based on device and or metric.
User Story
As a user I’d like to be able to write a script in my home directory (or some other persistent location) that returns a single value when run which would allow the reporting function to be setup to periodically call this script and ingest the result to plot over time. This would allow my to monitor things that I choose such as the temp of an HBA.
Restriction and validation could be put in place to keep the implementation simple and robust. e.g. “script must return a single numerical value”
Ideally there would also be some what of adding an alert based on the values collected to although I’m not sure if the alert functionality is tied to the report functionality in anyway or if it is a separate implementation?