Monitors per-disk SMART attributes on Synology NAS units via SNMP (SYNOLOGY-SMART-MIB), with configurable thresholds on the pre-failure counters.
Includes the following checks:

Per-disk SMART health: CRIT on any attribute with a status other than OK, except a resolved past failure, which stays OK by default and is always shown in the details.
Pre-failure counters: reallocated, pending, offline uncorrectable, reported uncorrectable, and UDMA CRC, default WARN at 1 and CRIT at 10.
Power-on hours, reported with a metric.

One service per physical disk, named by drive bay including DX-series expansion enclosures, with device-path naming as a fallback.
Version history:

1.1.0: resolved past failures no longer force CRIT, corrected drive bay naming on units with unordered tables or NVMe cache
1.0.1: author metadata update
1.0.0: initial release, per-disk SMART health, pre-failure counter thresholds with a dedicated ruleset, error counter and power-on hours metrics