downtimed(8)

downtimed8

System downtime monitoring and reporting daemon

Usage

downtimed [-D] [-d datadir] [-F] [-f timefmt] [-l log] [-p pidfile] [-S] [-s sleep]

downtimed -v

Overview

The downtimed daemon waits in the background, frequently updating a time stamp file on the disk. If the daemon is killed with a signal associated with a normal system shutdown procedure, it records the shutdown time on the disk.

When the daemon is restarted during the next boot process, it reports how long the system was down and whether it was properly shut down or crashed. The downtime report is output to the system log or to a specified log file. Also a record is appended to the downtime database.

Options

-D

Do not create nor update the downtime database.

-d datadir

The directory where the time stamp files as well as the downtime database are located. The default directory is determined at compile time.

-F

Do not call daemon(3) to fork(2) to background. Useful with modern system service managers such as systemd(8), launchd(8) and others.

-f timefmt

Specify the time and date format to use when reporting using strftime(3) syntax. The default is "%F %T".

-l log

Logging destination. If the argument contains a slash (/) it is interpreted to be a path name to a log file, which will be created if it does not exist already. Otherwise it is interpreted as a syslog facility name. The default logging destination is "daemon" which means that the messages are written to syslog with the daemon facility code.

-p pidfile

The location of the file which keeps track of the process ID of the running daemon process. The system default location is determined at compile time. May be disabled by specifying "none".

-S

Normally fsync(2) is performed after each update of the time stamp. This option disables the fsync(2). It reduces the load on the disk system but makes the downtime measurement less reliable.

-s sleep

Defines how long to sleep between each update of the on-disk time stamp file. More frequent updates result in more accurate downtime reporting in the case of a system crash. Less frequent updates decrease the amount of disk writes performed. The default is to sleep 15 seconds between each update. If you are using a flash memory based SSD or other disk which has limited amount of write cycles per block, it might be a good idea to set the sleep time to a higher value to prolong the lifetime of the storage device.

-v

Display the program version number, copyright message and the default settings.

Signals

SIGHUP

Close and re-open the output log. Use in case you want to rotate the log file.

SIGTERM and SIGINT

Terminate gracefully. These signals signify that a graceful system shutdown is in process.

Exit status

The daemon exits 0 on success, and >0 if an error occurs.

Releases

7 releases since Debian 7 (Wheezy); the text changed 2 times.

  1. 7
  2. 8
  3. 9–13

-F since Debian 9 (Stretch); -f since Debian 8 (Jessie).

Bugs

The reporting accuracy in case of a system crash depends on how often the time stamp is updated.

Finding out the system startup time is very operating system specific. If the program does not have specific code to support your operating system, it assumes that the system started when the daemon started.

Reporting is inaccurate if the system clock changes during system downtime or startup process. Daylight saving time changes have no effect as all calculations are done using UTC.

From downtimed 1.0-4 in Debian 13 (Trixie). Parsed by carta from the page's man source.