[mythtv-users] All's well with DataDirect again

Sun Apr 8 04:35:53 UTC 2007

On 04/07/2007 10:36 PM, Bruce Markey wrote:
> Michael T. Dean wrote:
>   
>> On 04/07/2007 07:08 PM, Kevin Atkinson wrote:
>>     
>>> On Sat, 7 Apr 2007, Michael T. Dean wrote:
>>>> On 04/07/2007 03:32 PM, Yeechang Lee wrote:
>>>>> We have reloaded the data. The Digital Stations that were running out
>>>>> of data on 4/8/2007 should now have a full schedule out.
>>>> And remember that mythfilldatabase will pick up these changes on its
>>>> daily runs--you don't need to force a refresh.  Tomorrow's data will be
>>>> refreshed today.  +2's day will be refreshed tomorrow, +3's on +2, and
>>>> so on...
>>>>
>>>> For those days that you had a significant amount of "No Data," they will
>>>> be refreshed on the next mythfilldatabase run.
>>> However, unless I misunderstand something today's data will not be picked 
>>> up.  Therefore you might want to run
>>>
>>> ...
>> If your mythfilldatabase ran early today and your next scheduled 
>> run (which is available from the backend status web page or the System 
>> Status screen in mythfrontend) isn't until tomorrow and you need to 
>> ensure the proper listings for Sunday, run mythfilldatabase without any 
>> "--refresh" arguments.
> The first thing to do right out of the gate is:
>
> $ mythfilldatabase --max-days 1
>
> This grabs today's data only and ought to be done ASAP. The
> rest isn't as urgent.
>
> The new data was not available until midday (past noon PDT)
> so any grab earlier than that did no good. In fact, they hadn't
> done their usual daily load within the previous 24 hours (a time
> which I won't specify ;) so a grab before noon PDT would have
> gotten the same bad data as the day before.
>
> Because myth grabs data for 'tomorrow' and any new days added
> two weeks out, it would be wise to reload all days to replace
> the damaged dozen or so days in between. Correcting these may
> help with conflict resolution when future showings are known.
> This can work both ways so a show could be Latered to resolve
> a conflict OR a show may currently be Latered to a showing
> which will no longer be shown (this would be bad and result in
> missing an episode).
>
> $ mythfilldatabase --refresh-all
>
> This should only be done when you know that the data in your
> database is incorrect, say, today for example.

OK, fine, feel free to run a --refresh-all if it makes you feel better.  
;)  But, all you users, before you do, *please* at least check your logs 
to see what happened on your normal run today.  (You do have 
mythfilldatabase set to log its results, right?  If not, please check 
your listings before forcing a refresh.)

Why?  Well, chances are Myth's normal update got a /bunch/ of data for 
you already.  The below is straight out of my log file from today's 
/scheduled/ daily run of mythfilldatabase (i.e. I did /not/ run it 
manually to force a refresh) using the /normal/ arguments (which, in my 
case is "--no-delete").  Notice that 13 days of data were refreshed:  
every day except "today" (which was actually yesterday--i.e. the last 
few hours of Saturday--when mfdb ran) was reloaded.

And, since I didn't care about "today" (I seldom record on Saturdays, 
and when I do, none of it is important), it means that the normal run 
got all the important stuff a "--refresh-all" would have gotten for me 
(only difference is that I didn't get Saturday's data).  Since Saturday 
is basically over (due to timezone, it's not quite, but...), my running 
a "--refresh-anything" would be a waste.

Thanks,
Mike

2007-04-07 17:10:02.813 Checking day @ offset 0, date: Sat Apr 7 2007
2007-04-07 17:10:02.835 Data is already present for Sat Apr 7 2007, skipping
...
2007-04-07 17:10:02.835 Checking day @ offset 1, date: Sun Apr 8 2007
2007-04-07 17:10:02.835 Refreshing data for Sun Apr 8 2007
...
2007-04-07 17:10:10.775 Checking day @ offset 2, date: Mon Apr 9 2007
2007-04-07 17:10:10.797 Data refresh needed because only 15 out of 30 
channels have at least one program listed for day @ offset 2 from 8PM - 
midnight.  Previous day had 25 channels with data in that time period.
...
2007-04-07 17:10:14.358 Checking day @ offset 3, date: Tue Apr 10 2007
2007-04-07 17:10:14.381 Data refresh needed because only 16 out of 30 
channels have at least one program listed for day @ offset 3 from 8PM - 
midnight.  Previous day had 24 channels with data in that time period.
...
2007-04-07 17:10:18.365 Checking day @ offset 4, date: Wed Apr 11 2007
2007-04-07 17:10:18.387 Data refresh needed because only 15 out of 30 
channels have at least one program listed for day @ offset 4 from 8PM - 
midnight.  Previous day had 25 channels with data in that time period.
...
2007-04-07 17:10:22.147 Checking day @ offset 5, date: Thu Apr 12 2007
2007-04-07 17:10:22.170 Data refresh needed because only 16 out of 30 
channels have at least one program listed for day @ offset 5 from 8PM - 
midnight.  Previous day had 24 channels with data in that time period.
...
2007-04-07 17:10:25.855 Checking day @ offset 6, date: Fri Apr 13 2007
2007-04-07 17:10:25.881 Data refresh needed because only 15 out of 30 
channels have at least one program listed for day @ offset 6 from 8PM - 
midnight.  Previous day had 25 channels with data in that time period.
...
2007-04-07 17:10:29.763 Checking day @ offset 7, date: Sat Apr 14 2007
2007-04-07 17:10:29.787 Data refresh needed because only 16 out of 30 
channels have at least one program listed for day @ offset 7 from 8PM - 
midnight.  Previous day had 24 channels with data in that time period.
...
2007-04-07 17:10:33.499 Checking day @ offset 8, date: Sun Apr 15 2007
2007-04-07 17:10:33.523 Data refresh needed because only 2 out of 30 
channels have at least one program listed for day @ offset 8 from 8PM - 
midnight.  Previous day had 25 channels with data in that time period.
...
2007-04-07 17:10:36.696 Checking day @ offset 9, date: Mon Apr 16 2007
2007-04-07 17:10:36.721 Data refresh needed because only 2 out of 30 
channels have at least one program listed for day @ offset 9 from 8PM - 
midnight.  Previous day had 25 channels with data in that time period.
...
2007-04-07 17:10:40.079 Checking day @ offset 10, date: Tue Apr 17 2007
2007-04-07 17:10:40.106 Data refresh needed because only 2 out of 30 
channels have at least one program listed for day @ offset 10 from 8PM - 
midnight.  Previous day had 24 channels with data in that time period.
...
2007-04-07 17:10:43.417 Checking day @ offset 11, date: Wed Apr 18 2007
2007-04-07 17:10:43.446 Data refresh needed because only 2 out of 30 
channels have at least one program listed for day @ offset 11 from 8PM - 
midnight.  Previous day had 25 channels with data in that time period.
...
2007-04-07 17:10:47.032 Checking day @ offset 12, date: Thu Apr 19 2007
2007-04-07 17:10:47.062 Data refresh needed because only 0 out of 30 
channels have at least one program listed for day @ offset 12 from 8PM - 
midnight.  Previous day had 24 channels with data in that time period.
...
2007-04-07 17:10:50.443 Checking day @ offset 13, date: Fri Apr 20 2007
2007-04-07 17:10:50.474 Data refresh needed because only 0 out of 30 
channels have at least one program listed for day @ offset 13 from 8PM - 
midnight.  Previous day had 25 channels with data in that time period.