Custom Query (1405 matches)

Currently, the -o on option enable the Automatic Offline Testing for every 4 hours. It is hard-coded constant.

I know the concept of Automatic Offline Testing is obsolete, some vendors doesn't implement such command at all, while others accept it, but consider it NOOP.

Despite of it, on the drives where implemented as non-void would be nice to have a way how to request different time.

I would like to introduce backward compatible option

-o on[,N]

where N specify the time in the form <number>[<unit>]

Unit may be 's' for seconds, 'm' for minutes, 'h' for hours. Seconds assumed when not present.

Missing N is assumed to be 4h to maintain backward compatibility.

The time is rounded to nearest lower available value (with exception that 1-4s are rounded to 5s). Such value is reported back to user in place of original message

SMART Automatic Offline Testing Enabled every four hours.

Some recent (e.g. Samsung) drives report errors only in the Extended Comprehensive SMART error log. The Summary SMART error log can be read but is always empty. In this case, smartd ('-l error' directive) will never report an error.

Smartd should check the error counters from both error logs.

By default smartd sends one mail for each type of disk problem detected ('-M once' directive). If '--savestates' is enabled, the mail state is also stored such that a mail is never send again even if smartd is restarted.

Workaround: specify '-M daily'.