[mythtv-users] Master backend thinks manually awoken slave is still sleeping
mythtv at theseekerr.com
Sun Feb 14 05:53:49 UTC 2010
On Sun, Feb 14, 2010 at 4:39 PM, Chris Pinkham <cpinkham at bc2va.org> wrote:
> * On Sun Feb 14, 2010 at 12:06:44PM +1100, Christopher Kerr wrote:
>> A new issue: having configured my slave backend to power down when not
>> needed, I find it is difficult to use this backend for LiveTV.
> LiveTV ugggghhh....
>> I understand that by design, the master will not wake a backend to use
>> a tuner for liveTV. However, if I choose to power the backend on
>> myself, in order to make the tuner available, the master still
>> considers the tuners as "asleep" and will not use them.
> I'm not sure how/why this is happening. When a slave connects, the master
> will set it's "sleep status" to either 'undefined' or 'awake'. Undefined
> is used if the slave does not have a SleepCommand configured, Awake is
> used if the slave does have a SleepCommand.
> So, your slave backend's sleep status should be changed to either Awake or
> Undefined when it connects to the master whether it is started manually or
> automatically by the master.
Is there a level of logging verbosity that will let me see those
changes being made?
> Are you assuming that the slave is still asleep because you can't use
> it's tuner for LiveTV or are you looking at the master backend's status
> webpage or the frontend's status screen?
The frontend's status screen. The frontend is on the same machine as
the slave backend.
>> If I schedule a recording, so the master sends the wake signal and
>> considers the slave to be awake, LiveTV on the other tuner works as
> If you are tracking trunk, I did fix one bug I noticed when looking into
> this, but don't think it was related, but if you want to svn update and give
> the current version a try, you can. If you're on 0.22 and compile yourself,
> you can try a 1-line change to programs/mythbackend/mainserver.cpp if you want.
> If you're on 0.22-fixes, in mainserver.cpp, look for a line like this:
> if (!elink->IsWaking())
> Change that to look like this:
> if (! (elink->IsWaking() || elink->IsAsleep()))
> That will make it match trunk. I don't think that will have much to do with
> what you're reporting, but it's worth testing.
I'm running the a version of trunk about 3 days old, should I try updating?
- Chris Kerr
More information about the mythtv-users