[vdr] Re: Feature Request: restart on UPT

[Klaus Schmidinger <Klaus.Schmidinger@cadsoft.de>]

Martin Schoenbeck wrote:
> 
> Hi Klaus,
> 
> Martin Schoenbeck schrieb:
> 
> > I just did some grep on my syslog. During the last month I had 7 UPT and
> > 27 VDSB ,-) The UPT where on four days (25th, 26th, 28th and 29th of
> > January) with four intermixing VDSB. Perhaps it's due to comparable
> > reasons, but nobody was annoyed by VDSB by now, because it was cured by
> > restart automatically. May be the locking is a little bit faster with
> > good weather and then an UPT occurs, otherwise an VDSB.
> >
> > I didn't get any dots in my log up to now. But with an average of one
> > restart a day, something should happen soon.
> 
> It was not obviously better here. I had four UPT and seven VDSB here
> within seven days. This seems to be even a bit more, but I think, this
> is within measurement precision.
> 
> But I have two observations, which may be interesting. I've also
> patched, that vdr does a restart as soon as an UPT is seen. But now
> there were up to three errors and 'initiation emergency exit' until vdr
> really restart.

The emergency exit is initiated from within a thread, while the actual
exit is done in the main loop. So there may well be several 'initiation emergency exit'
messages.

> And one UPT arose on 'ARD', which is also the channel
> for setting the clock. Between starting the recorder thread and the
> first UPT there were 17 'System Time' 'Local Time' entries in syslog
> switching the clock back and forth a few seconds.

I don't think that this has anything to do with the UPT problem.

The next developer version of VDR will initiate an emergency exit
after 10 UPT errors per recording.

Klaus


-- 
Info:
To unsubscribe send a mail to ecartis@linuxtv.org with "unsubscribe vdr" as subject.