Adaptive refresh rate #3
Description
Refreshing the data and especially sending a HTTP request to healthchecks every 15 seconds seems to be a bit often for background operation. This could be adapted depending on the following conditions:
- borgmatic status (running, error, inactive)
- menu (and log output submenu) open or not
- interesting events (systemctl state/condition event, healthchecks ping/deadline)
- started via button
Still, the durations displayed in the menu could be refreshed every second if the menu is open without requiring polling of the data sources. Healthchecks probably only needs updates around the "interesting events" (or between different borgmatic app backup steps), while we need to poll systemd regularly to find out when these happen. This could boil down to the following 4 config variables: {ACTIVE,INACTIVE}_{SYSTEM,BORGMATIC}_POLL
Further ideas:
- Also, we should respect the system data saving mode (which should be mutually exclusive with backup operations).
- We could also incorporate better exception handling and complain once in a while if things keep failing.
- The log output should display the time field, maybe grouped by minute.
- Optional debug menu with information on last system / healthcheck polls and option for restarting borgmacator.