[mythtv-users] recording schedules cleanup, automatic removal of "record once" rules?

[Michael T. Dean]

On 11/08/2014 11:52 AM, Steven Adeff wrote:
> On Tue, Nov 4, 2014 at 1:38 PM, Michael T. Dean wrote:
>> On 11/04/2014 01:06 PM, Steven Adeff wrote:
>>> If I understand correctly based on the nomenclature used,
>>> "Record only this showing"
>>> "Find and record one showing of this title"
>>>
>>> at some point, MythTV will record that showing, or, it won't, and the
>>> date will pass, and in both cases, that recording rule will never
>>> "hit" again.
>>>
>>> If this is the case, then is there any reason those rules cannot be
>>> deleted? Is there any reason they cannot be deleted "automatically"
>>> after either of those two situations occur?
>>
>> This showing rules are deleted automatically (within a couple days) after
>> the showing's end time. Find rules are cleaned up automatically once a
>> successful recording is completed--which means if your recording is marked
>> as damaged, for example, they won't be cleaned up (and will cause a
>> recording for a future showing if you haven't deleted the rule manually in
>> the meantime).
> Mike, let's say those rules are not being cleaned up automatically, is
> there a scenario when this would happen that this system might be
> triggering so that they remain?

If your master backend is running properly and jobs are running 
properly, then rules should be cleaned up--unless you have (some pretty 
major) database corruption (probably schema corruption, unless someone 
has been editing record-related data in the database directly).  That 
said, a far more likely scenario is that you've got a lot of shows 
marked as damaged (due to reception or capture device/recording errors), 
so the find rules are sticking around.

The queries are generally time-based, so it may be possible for time 
issues on the system to cause problems.  Possibly if your database is 
lacking time zone data?

Really, though, I'm just guessing.  I don't have any information to make 
any kind of educated guess.  There's not really any logging in the part 
that cleans up the database (unless you get database errors) because it 
just works (so we haven't even added any debug-level output for even any 
specific logging area).  You'd have to run the master backend with -v db 
and watch to see if it even runs the queries (look for "DELETE FROM 
record").

Mike