[mythtv] Scheduler needs table keys?
bjm at lvcm.com
Sun Jan 28 07:34:26 UTC 2007
Chris Pinkham wrote:
> So from this quick test, you can see how drastic the differences can be
> in different people's scheduler runtimes based solely on their types of
> scheduled recordings and the number of each.
Verified. With 122 channels and 13 days of data for 42718
program entries, I swapped type 3 and type 4 back and forth
for 70 test rules and ran mythbackend --resched from another
window. The match phase times as kAllRecord was consistently
~10.70 sec whereas kChannelRecord matched in ~0.87.
Note that there are all sorts of inequities here most importantly
that this is a full reschedule like starting the MBE or running
mfdb rather than a change to one rule which is more typical
during normal operation.
I'd always favored All for simplicity but checking a few
hundred programs on a channel is much quicker that searching
thirty or forty thousand programs.
More information about the mythtv-dev