v1.4.5
Fixed: Package and Docker Metrics on Slow Devices
This release addresses an issue affecting slower systems such as Raspberry Pis and other SBCs. Package and Docker metric collection could exceed the short execution limits, causing pending package updates to be incorrectly reported as 0 with an empty package list.
Changes
- Package update and Docker statistics collection are now separated from the regular monitoring cycle.
- Fast system metrics can continue to update every 30 seconds without waiting for slower commands.
- Package metrics are collected every 12 hours by default.
- Docker metrics are collected every 30 minutes by default.
- Slow collectors now use a separate 180-second timeout.
- Collection intervals and the slow-collector timeout are configurable.
- Timed-out collections preserve the last valid values instead of replacing them with zero or empty data.
- A manual refresh also triggers package and Docker collection.
- New diagnostic sensors report collection duration and errors separately for package and Docker metrics.
Full Changelog: v1.4.4...v1.4.5