[mythtv-users] Missing EPG on a number of channels

[Robert]

On 23-01-27 23:35:48 CET, Klaas de Waal wrote:
> On Fri, 27 Jan 2023 at 20:23, Robert <2201mythtvml at rainbow.in-berlin.de>
> wrote:
> 
> > Dear MythTV users,
> >
> > I'm missing EPG for a number of channels. Unlike one year ago, this time
> > it is not on the provider's side: two neighbours have the EPG on the
> > channels (on their normal TV sets) where I haven't got it in MythTV.
> > I did a full rescan, but this did not fix the problem.
> > The affected channels share a small number of mplexid values in the
> > database, one these multiplexes no channel seems to have EPG.
> > useonairguide is 1 for all of them.
> >
> > What can be the reason for the missing EPG on just a few multiplexes?
> > Why doesn't a rescan fix this?
> > How can I get the EPG back? Delete the affected channels and do a rescan?
> >
> > I have a quadHD card connected to some kind of DVB-C that comes via FTTH
> > (fiber), in Germany.
> > I'm running version 1:31.0+fixes20220227.git7e4ce1ba98-dmo0+deb11u1 from
> > http://ftp.uni-kl.de/debian-multimedia/.
> >
> > Setting the verbose to eit and loglevel to debug did not give me any
> > clue. mplexid is one of the ones lacking EPG data.
[...]
> 
> Sometimes there is guide information sent as private data which can only be
> used by set-top-boxes from the provider.
> MythTV can only use guide data that is coded according to the DVB standards.
> However, if a standard TV without a set-top-box can show the guide then I
> expect that MythTV should also be able to do that also.
> 
> I suggest first upgrading to the latest fixes/32.

Last weekend, without upgrading, I deleted all channels and rescanned.
This brought back EPG in some of the channels. Weirdly, both Mythweb and
the backend's own Web interface seemed confused. For certain channels,
Mythweb did not show its program, but it showed upcoming recordings!
I wondered whether all this was due to the old version, so this weekend,
I switched to deb-multimedia.org's bullseye-backports. It provides a
version 1:32.0+fixes20230124.gitf2ce7e33aa-dmo0~bpo11+1.

> Then with mythtv-setup do the following
> - delete all channels
> - do a full scan with the following log options "-v
> channel,chanscan,eit,siparser --loglevel=debug"
> - capture all output of the full scan, e.g. with "mythtv-setup -v
> channel,chanscan,eit,siparser --loglevel=debug 2>&1 | tee
> this_is_my_logfile.log"

I did this, but this time I deselected "Remove duplicates". My source is
supposed to be DVB-S received by the Internet provider who feeds it to
us via fiber to the home.
I chose to insert the channels manually. This meant having to enter all
the channels' numbers again, quite a bit of work. Interestingly, many
channels appeared twice or more, even though it's supposed to be
not DVB-T2.

Again, EPG is missing on some channels, at least at their first
appearance which I gave the old number.
But - EPG is present on another channel with the same name which I gave
a higher number.
One example, I hope the query makes sense:

MariaDB [mythconverg]> select chanid, visible, useonairguide, deleted, count(*) as eit_count, from_unixtime(max(endtime)) as maxend from channel left join eit_cache using (chanid) where name='ONE HD' group by chanid order by maxend;
+--------+---------+---------------+---------------------+-----------+---------------------+
| chanid | visible | useonairguide | deleted             | eit_count | maxend              |
+--------+---------+---------------+---------------------+-----------+---------------------+
|  10005 |       1 |             1 | 2023-01-29 17:14:37 |         1 | NULL                |
|  16376 |      -1 |             0 | NULL                |         1 | NULL                |
|  60005 |      -1 |             0 | NULL                |         1 | NULL                |
| 501429 |       1 |             1 | NULL                |         1 | NULL                |
|  20354 |       1 |             1 | 2023-02-05 10:21:04 |       681 | 2023-03-01 02:05:00 |
|  10405 |       1 |             1 | NULL                |       796 | 2023-03-05 01:10:00 |
+--------+---------+---------------+---------------------+-----------+---------------------+

If I understand this correctly, I've got two visible non-deleted
channels, one with EIT data (10405), the other without (501429).

Another example:

MariaDB [mythconverg]> select chanid, visible, useonairguide, deleted, count(*) as eit_count, from_unixtime(max(endtime)) as maxend from channel left join eit_cache using (chanid) where name='TLC' group by chanid order by maxend;
+--------+---------+---------------+---------------------+-----------+---------------------+
| chanid | visible | useonairguide | deleted             | eit_count | maxend              |
+--------+---------+---------------+---------------------+-----------+---------------------+
|  10148 |       1 |             1 | 2023-01-29 17:14:37 |         1 | NULL                |
|  20397 |       1 |             1 | 2023-02-05 10:21:04 |         1 | NULL                |
| 501425 |       1 |             1 | NULL                |         1 | NULL                |
|  10548 |       1 |             1 | NULL                |       128 | 2023-02-09 01:05:00 |
+--------+---------+---------------+---------------------+-----------+---------------------+

Is this normal to have multiple channel records for the same channel?

> Please create a ticket for this issue on
> https://github.com/MythTV/mythtv/issues and then attach the log file to
> that ticket.

I'm not sure what issue to report. 
Should I get rid of the extra non-deleted channels first? How?

> Then when you run mythbackend use the following log options: "-v
> channel,eit,siparser --loglevel=debug", capture the output for a few days
> and post that also.
> This will also be a fairly big file.

It is currently running that way, in case it's needed.

Thanks,
Robert