[systemd-devel] [EXT] Re: Understanding the effect of AccuracySec=
Windl, Ulrich
u.windl at ukr.de
Mon Aug 19 05:55:28 UTC 2024
> -----Original Message-----
> From: Andrei Borzenkov <arvidjaar at gmail.com>
> Sent: Saturday, August 17, 2024 8:22 AM
> To: Windl, Ulrich <u.windl at ukr.de>; systemd-devel at lists.freedesktop.org
> Subject: [EXT] Re: [systemd-devel] Understanding the effect of
> AccuracySec=
>
> On 16.08.2024 12:43, Windl, Ulrich wrote:
> > Hi!
> >
> > I have defined a time using
> >
> > [Timer]
> > OnCalendar=daily
> > AccuracySec=6h
> > Persistent=true
> >
> > And the idea was to run the unit daily, but it's rather unimportant when,
> just nout in prime hours.
> >
> > The man page says:
> >
> > AccuracySec=
> > Specify the accuracy the timer shall elapse with. Defaults to 1min.
> > The timer is scheduled to elapse within a time window starting with
> > the time specified in OnCalendar=, OnActiveSec=, OnBootSec=,
> > OnStartupSec=, OnUnitActiveSec= or OnUnitInactiveSec= and ending
> > the time configured with AccuracySec= later. Within this time
> > window, the expiry time will be placed at a host-specific,
> > randomized but stable position that is synchronized between all
> > local timer units. This is done in order to distribute the wake-up
> > time in networked installations, as well as optimizing power
> > consumption to suppress unnecessary CPU wake-ups. To get best
> > accuracy, set this option to 1us. Note that the timer is still
> > subject to the timer slack configured via systemd-system.conf(5)'s
> > TimerSlackNSec= setting. See prctl(2) for details. To optimize
> > power consumption, make sure to set this value as high as possible
> > and as low as necessary.
> >
> > So I expected that multiple instances of the timer would be "spread", but
> instead I see all instances started at 00:00:01.
>
> It sounds like you want RandomizedDelaySec, not AccuracySec.
[Windl, Ulrich]
Despite of the fact that my version of systemd does not know that setting, can you explain what the effect of AccuracySec= really is?
[...]
More information about the systemd-devel
mailing list