Tvheadend

If more services per channel are configured (usually on different tuners), there is already some priority logic applied, in which order the services are tried.

The method service_get_prio (https://github.com/andoma/tvheadend/blob/master/src/service.c#L241) could be changed to add some configurable weight per adapter.

I guess there needs to be some thinking about when trying to steal a transponder. For the moment, the simple solution would be to apply this adapter priority only on idle adapters.

I'm gonna try this within the next days.

Well, an initial version for proof of concept is ready, working but not really final.

The issue I'm running into is that when deciding the order of the services (and therefore the tuners) which are tested, the quality of the services is the first criteria.

For now, I'm disabling quality mechanism if an extra weight is configured on the adapters, since may mean that the user doesn't want to have automatic behavior but give the order manually.

Sorry to drag an older feature request up, but I've now hit the same problem.
I too have DVB-T (Terrestrial) and DVB-S (Satellite), and both carry the same channels. However the input from DVB-T is lower quality (not sure if this is the decoder I have, bad aerial, etc), BUT the Multiplexes always show 100% quality. The DVB-S is currently also showing quality at 100%, but it seems the DVB-T card is being preferred over the DVB-S cards. I think your weight technique may well help this out, but if not can you advise on best options here?
I'm currently using the git head master, so I believe I'm on 2.12.99 but happy to test, etc, if you can advise?

Thanks Martin. I've patched and got that running now.

Just to note what I've done, I've added 50 to the DVB-T card and left the rest with a 0 "Extra Priority".
Will do some further testing the next couple of days. Thanks for the patch!

Applied the patch above last night and everything works perfectly, assigned the DVB-S a weight of 50 and the DVB-T a weight of 0 and tested the channels that are available in HD via the DVB-T and SD via the DVB-S and the DVB-T HD channel was selected. Exactly what i wanted.

Can this be merged into the current 2.13 git? This is working fine here and have tested various functions and all appears fine.

I'll update my repo and place a pull request within the next couple of hours.

Martin Mrvka wrote:

I'll update my repo and place a pull request within the next couple of hours.

Pull request added: https://github.com/andoma/tvheadend/pull/94

After updating to the latest revision (4ef72fb276), I get a SEGFAULT when tuning to a channel, which contains two DVB-S and an IPTV configuration.

Jun 22 11:07:08 localhost tvheadend[12235]: CRASH: Signal: 11 in PRG: /usr/local/bin/tvheadend (2.99.44.g4ef72) [a1249437a3332ce32b816b661495fd11be08bbf6] CWD: /
Jun 22 11:07:08 localhost tvheadend[12235]: CRASH: Fault address 0x118 (Address not mapped)
Jun 22 11:07:08 localhost tvheadend[12235]: CRASH: Loaded libraries: /lib/x86_64-linux-gnu/librt.so.1 /lib/x86_64-linux-gnu/libdl.so.2 /usr/lib/x86_64-linux-gnu/libavahi-common.so.3 /usr/lib/x86_64-linux-gnu/libavahi-client.so.3 /lib/x86_64-linux-gnu/libcrypto.so.1.0.0 /lib/x86_64-linux-gnu/libpthread.so.0 /lib/x86_64-linux-gnu/libc.so.6 /lib64/ld-linux-x86-64.so.2 /lib/x86_64-linux-gnu/libdbus-1.so.3 /lib/x86_64-linux-gnu/libz.so.1
Jun 22 11:07:08 localhost tvheadend[12235]: CRASH: Register dump [23]: 0000000000000000 00000000020f4f10 0000000000000001 0000000000000000 00007f09c97f9110 0000000000000001 00007f09c97f9070 0000000000000000 0000000000000007 0000000000455c95 00007f09c97f90f0 0000000001d1d4f0 000000000044f3c0 0000000000000000 00007f09c97f9110 00007f09c97f9060 0000000000412a5b 0000000000010202 0000000000000033 0000000000000004 000000000000000e fffffffe7ffbfa17 0000000000000118
Jun 22 11:07:08 localhost tvheadend[12235]: CRASH: STACKTRACE
Jun 22 11:07:08 localhost tvheadend[12235]: CRASH: /usr/src/tvheadend/src/trap.c:139 0x420b9b
Jun 22 11:07:08 localhost tvheadend[12235]: CRASH: ??:0 0x7f09db186cb0
Jun 22 11:07:08 localhost tvheadend[12235]: CRASH: /usr/src/tvheadend/src/service.c:346 0x412a5b
Jun 22 11:07:08 localhost tvheadend[12235]: CRASH: /usr/src/tvheadend/src/subscriptions.c:173 0x411b98
Jun 22 11:07:08 localhost tvheadend[12235]: CRASH: /usr/src/tvheadend/src/subscriptions.c:326 0x411d87
Jun 22 11:07:08 localhost tvheadend[12235]: CRASH: /usr/src/tvheadend/src/webui/webui.c:411 0x434574
Jun 22 11:07:08 localhost tvheadend[12235]: CRASH: /usr/src/tvheadend/src/http.c:344 0x407e7d
Jun 22 11:07:09 localhost tvheadend[12235]: CRASH: /usr/src/tvheadend/src/http.c:375 0x408243
Jun 22 11:07:09 localhost tvheadend[12235]: CRASH: /usr/src/tvheadend/src/http.c:455 0x4083e2
Jun 22 11:07:09 localhost tvheadend[12235]: CRASH: /usr/src/tvheadend/src/http.c:749 0x4087c0
Jun 22 11:07:09 localhost tvheadend[12235]: CRASH: /usr/src/tvheadend/src/tcp.c:397 0x406c6d
Jun 22 11:07:09 localhost tvheadend[12235]: CRASH: ??:0 0x7f09db17ee9a
Jun 22 11:07:09 localhost tvheadend[12235]: CRASH: clone+0x6d  (/lib/x86_64-linux-gnu/libc.so.6)
Jun 22 11:07:09 localhost kernel: [2231907.860406] tvheadend[13904]: segfault at 118 ip 0000000000412a5b sp 00007f09c97f9060 error 4 in tvheadend[400000+90000]

Thanks for the bugreport. I'll look at it asap! I may have put DVB code where it shouldn't be....

I've replaced DVB dependend code, however I didn't get any working IPTV stream (tried hard with mumudvb) in order to test.

@Bianco
Would you mind to checkout my branch https://github.com/mmrvka/tvheadend/tree/dvb_adapter_priority and try if it works for you?

I will try this tomorrow. Thanks for your fast response.

Finally I got IPTV working and was able to test. BTW it wasn't the setup of mumudvb which caused problems - it was simply that I didn't knew how to add a multicast address in tvheadend (forgot the @ infront of @239.255.0.0).

My test with two DVB adapters and one IPTV stream associated to the same channel have been successful.

Pull request opened: https://github.com/andoma/tvheadend/pull/102

Finally I've been able to test yout patch, and it works as expected. Thanks again.