Marginal signal on HDHR causes excessive error logging

[Marc Randolph <mrand@…>]

With the best digital signal he is able to achieve, a downstream user is encountering hard drive thrashing that appears to be due to excessive log entries (see below... at least four per second)

2009-10-31 23:52:54.559 HDHRSH(10172A63-0): RunTS(): data_length = 48692 remainder = 188
2009-10-31 23:52:54.594 HDHRSH(10172A63-0): RunTS(): data_length = 57904 remainder = 188
2009-10-31 23:52:54.628 HDHRSH(10172A63-0): RunTS(): data_length = 53956 remainder = 188
2009-10-31 23:52:55.264 HDHRSH(10172A63-0): RunTS(): data_length = 3948 remainder = 188

Perhaps a fix could be as simple as only logging repeated attempts such as this once every 2 (or even 5) seconds?

[callmegar@…]

I'm suffering this problem too, myth-backend is logging thousands of lines with messages from hdhrstreaming.cpp (this RunTS() message) and dtvsignalmonitor.cpp (Channel not found messages) even when reception is mostly good and live TV is working fine.

[mythtv@…]

Seeing this here too on 0.22 fixes. The annoying part is it happens even when not using the frontend, and with nothing recording on the HDHR. Mythbackend.log has grown to over 8MB since the last logrotate.

[robertm]

In trunk at least, this logging only occurs with -v record on the backend. If you don't want to see verbose logging of recording messages, simply use default logging.