check_ide_smart: introduces risk of data loss

[waja]

In Debian we got reported this issue: https://bugs.debian.org/1010038

I've been relying on check_ide_smart for some time to monitor my HDDs and up to
now there has never been a malfunction reported, with all disks being reported
as passing all tests and in the OK status. While setting up a new server I've
found two issues with this plugin that I think should be enough to mark it as
deprecated, both in Debian and in upstream:

• It just doesn't recognize drives connected through an HBA, even though
  smartctl does. This is a known limitation of the plugin.

• When using a more up to date alternative, I've found that some of the HDDs
  that check_ide_smart was reporting as OK have in fact reached some pre-failure
  conditions and are worth replacing. This isn't related to the hardware
  interface as those drives are in production and have been monitored with
  check_ide_smart for quite some time, I've just changed the monitoring plugin.

So, as check_ide_smart isn't properly reporting HDD status (or, at least,
failing to report early warnings) I believe keeping it available as is
introduces the risk for a potential data loss. There are alternatives
available, so I suggest to consider substituting the plugin instead of updating
it. The best I've found is Kurt Yoder and Claudio Kuenzler's check_smart[1][2],
but I'm not sure of the licensing status (the code shows that "This script was
initially created under contract for the US Government and is therefore Public
Domain", but there's no specific license mentioned).

Regards.

[1]https://www.claudiokuenzler.com/monitoring-plugins/check_smart.php
[2]https://exchange.nagios.org/directory/Plugins/System-Metrics/Storage-
Subsystem/check_smart-2Epl--2D-physical-drive-monitoring/details