Gather storage drives health status

[mokrinsky]

PR title kinda says for itself, also explained in Readme.

In short, sole reason for me to look for the idrac_exporter was the need of healthchecking A LOT OF DRIVES. But there was no such feature, so - casually coded.

Few things i have to mention just in case:

1. I did not implement entire data models for what I added, i just used some fields that looked useful to me.
2. All Dell servers I have available are equipped with just one RAID controller. I'm neither aware if there are any with several controllers nor I have them, so this PR will gather data from the first controller only.

[mrlhansen]

Hi @mokrinsky

Thanks for the PR it looks good! I will find some time to test it before merging.

[mrlhansen]

Hi @mokrinsky

It seems to work fine across multiple machines and vendors. There are a couple of things I would change.

• Change metric name to idrac_drive_health to better reflect what the value represents.
• I would also add status as a label for the metric. In case the status is not OK/Warning/Critical the state can be read from the label value.
• Add a loop over controllers in RefreshStorage(). For NVMe in particular, each drive has its own controller.

You are welcome to make the changes if you want, otherwise I will make the adjustments after merging.

[mokrinsky]

Hey! Thanks for your review.

• I renamed metric.
• I actually don't think it's necessary cause in Redfish specification there are only three statuses possible, but for the sake of possible weirdness in Dell hardware I regularly meet... Made this as well.
• It would be great if you'll adjust it yourself cause, like I said, i got no servers with such configuration so I cannot test it myself.