Project

General

Profile

Feature #1464

Retries of bad tuner should cycle across all identically configured available tuners until one works

Added by Richard Lloyd over 9 years ago. Updated over 9 years ago.

Status:
Rejected
Priority:
Normal
Assignee:
-
Category:
DVB
Target version:
-
Start date:
2012-12-22
Due date:
% Done:

0%

Estimated time:

Description

My quad tuner satellite card (TBS6984) will very occasionally drop the signal strength across most of the transponder frequencies on one tuner below the level at which tvheadend can record from. The other 3 tuners are usually OK at the time and sometimes it can be fixed by rebooting the PC. It could be a driver bug, but I am on the latest TBS driver for Linux (Ubuntu 12.04.1).

This drop results in a "No input detected" error from tvheadend as you'd expect and tvheadend does do retries continuously, but only on the same tuner that had the near-zero signal strength! I have all 4 tuners identically configured (same multiplexes, SIDs, services, channels etc) and I think tvheadend should cycle through each tuner on the system that is "identical" to the one with the signal problem on it (skipping ones that are already being recording from or actually overriding viewing streams if there's nothing else left).

As it stands, retrying just one tuner that's got a problem when three other identical ones are available and working is quite frustrating :-( You can see it struggling in the Active Subscriptions window - my only fix was to dive into the dodgy tuner's config, find the multiplex frequency for the channel I was trying to record, sort multiplexes in frequency order and tick off the particular multiplex on the problematic tuner. Only then did tvheadend decide to move to the second tuner and start recording the channel. This worked, but I shouldn't have to do all that manually surely?

History

#1

Updated by Adam Sutton over 9 years ago

  • Status changed from New to Rejected

Can you please resubmit as bug report with a full debug log. This is the way tvh is already designed to work and reading the code it looks right to me, but I don't think your the first to mention this so it's possible something got broken along the way.

Adam

Also available in: Atom PDF