[mythtv-users] Events at the beginning of a recording
Klaas de Waal
klaas.de.waal at gmail.com
Mon Feb 6 21:27:19 UTC 2023
On Mon, 6 Feb 2023 at 14:08, Paul Harrison <mythtv at mythqml.net> wrote:
> On 06/02/2023 11:17, Jan Ceuleers wrote:
> > On 05/02/2023 23:13, Klaas de Waal wrote:
> >> I think ideally you have to have WakeUpThreshold >= global
> pre-roll +
> >> 120. So with your 250 seconds pre-roll I would try something like
> >> 370 or
> >> 400 for WakeUpThreshold. That should give you about 120 seconds
> >> events if I understand the problem properly.
> >> Yes, I share your understanding of the logic but the HandleWakeSlave is
> >> not called early enough.
> >> My logs show most of the time only two REC_PENDING events, one on 60
> >> seconds and one on 30 seconds.
> >> I think I found where this comes from but I as this is the scheduler I
> >> need to be very careful....
> > I have changed the WakeUpThreshold to 600 (just to be on the safe side),
> > and I still only get two REC_PENDING events (at 60 and 29 seconds), same
> > as Klaas reports.
> > Cheers, Jan
> OK thanks for testing.
> The problem must be further up the code in the main scheduler loop which
> is not calling HandleWakeSlave early or often enough for the code that
> sends the REC_PENDING events to work properly. When it's working
> properly it should be sending events at approximately 120, 90, 60 and 30
> seconds before recording starts.
> There is the constant kSleepCheck = 300 and this is the maximum number of
seconds to wait between scheduler runs.
See commit 81dffd5ab6becb75f9e004ffe771519d23a58d87
Lawrence Rust authored and Paul Harrison committed on Jul 15, 2014
where this was changed from a hardcoded 300 to the constant.
When kSleepCheck is changed to 30 seconds then there is always one
in the first time window, from 120 to 90 seconds.
Paul, do you think that it is a problem for the scheduler to wake up every
> Klaas I wonder if we should be setting the default for WakeUpThresold
> to pre-roll + 120 instead of 5 minutes if that setting is not present in
> the DB?
I think that this is a good idea because if we do not do this then there is
on the RecordPreRoll value that is both unwanted and difficult to explain
to the casual user.
And there is no GUI option to change the WakeUpThreshold as far as I know.
In the slightly longer run we may consider separating WakeUpThreshold and
the WakeUpThreshold is about starting slave backends and that is not
necessarily related to REC_PENDING events.
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the mythtv-users