[Usability] Restart on resume (2 Viewers)

[mm1352000]

Good spotting
^^ that kind of random delay could be IPv6

 

[Requiem]

Okay, thanks, you are faster in tracking than me in generating logs and changing options.
Here is a third log, this time with server logs. I'll do a fourth one with watchdog olption #1 after this.

What was changed?
1. Used UNC paths was i bit tricky as the server uses a RAM-Disk for timeshifting, but adding a "net share" to the ramdisk startup script did the trick). Please ignore server error logs about not finding the timeshifting path, this is due to the ramdisk being initialized after the startup of TVService, the path is there a few seconds after the log entries).
2. Disabled MPSync, FritzBox-Manager and IFC plugins
3. Disabled Dynamic Refesh Rate
4. Changed from WASAPI to BASS
5. Disabled Startup with Windows
6. Resume WORKED this time (switched to standby with a remote during liveTV running, got back to the main menu).

Yes, my home network is IPv6 capable. I also get a "no connection to internet" from the WOrldWeather plugin after resume (which is fine a few seconds later), and this might also give a hint why the FritzBox manager is not capable to resolve the Router's hostname. Just wondering how to approach this one.

 

[Requiem]

Here is a set with watchdog option 1, resuming worked (gut to main menu, standby triggered on fullscreen live TV)

 

[HomeY]

OK, so in general we have good progress and a working resume situation when you disable most (all) plugins. That's a good thing
Based on your testing & results it looks like you can trigger the issue pretty easy, and no long standby is needed. We haven't seen any corrupted plugin databases so far, but based on the previous logs, i suspect the errors that MPSync throws to be the cause of corrupted databases.

Just wondering how to approach this one.

put the IP address & hostname in your hosts file.

I also get a "no connection to internet" from the WOrldWeather plugin after resume

That's normal since WorldWeather checks for updates after resume, and LAN needs a few seconds to init (i see it every day, but 1-2 seconds later all is working fine).

I'll go look @ the logs now.

 

[Requiem]

put the IP address & hostname in your hosts file.

You think about the routers IPv4 hostname, so like adding
192.xxx.xxx.xxx fritz.box

into the hosts file?
IPv6 works (if you specify the zone) with the hosts file, but it's a bit dirty....

 

[HomeY]

You think about the routers IPv4 hostname, so like adding
192.xxx.xxx.xxx fritz.box
into the hosts file?

Correct.

IPv6 works (if you specify the zone) with the hosts file, but it's a bit dirty....

Yes, when you add the Zone ID, it should also work with IPv6.
Are you using IPv6 on your Local network?

So the logs... MediaPortalLogs_11_06_14__15_01 is a bit tricky because it's hard to see when the system came out of standby. Did you resume it 1 minute after sending into standby?

Assuming you did, then resume seems fine:

[2014-06-11 14:53:09,727] [Log    ] [MPMain   ] [INFO ] - Main: OnResumeSuspend - Done

Then TVHome connects:

[2014-06-11 14:53:09,747] [Log    ] [TvClient-TvHome: HeartBeat transmitter thread] [INFO ] - TVHome: OnRemotingConnected, recovered from a disconnection

and then you let the system idle.

About 5 minutes later there is activity and the TvServer starts to throw errors:

[2014-06-11 14:58:41,427] [Log    ] [TvClient-TvHome: HeartBeat transmitter thread] [ERROR] - TvServer: Failed to execute remote TV server operation (IsAnyCardRecording) The operation has timed out after 3000.
[2014-06-11 14:58:45,437] [Log    ] [TvClient-TvHome: HeartBeat transmitter thread] [ERROR] - TvServer: Failed to execute remote TV server operation (IsAnyCardRecording) Der Objektverweis wurde nicht auf eine Objektinstanz festgelegt.
[2014-06-11 14:59:06,897] [Log    ] [MPMain   ] [ERROR] - TvServer: Failed to execute remote TV server operation (IsTimeToRecord.1) Ein Verbindungsversuch ist fehlgeschlagen, da die Gegenstelle nach einer bestimmten Zeitspanne nicht richtig reagiert hat, oder die hergestellte Verbindung war fehlerhaft, da der verbundene Host nicht reagiert hat 192.168.0.110:31456

but... those errors might be coming from another Suspend:

[2014-06-11 14:59:39,597] [Log    ] [MPMain   ] [INFO ] - TVHome.WndProc(): Windows is suspending
.....
[2014-06-11 15:00:30,672] [Log    ] [MPMain   ] [INFO ] - Main: OnSuspend - Done

which was resumed instantly?

[2014-06-11 15:00:39,913] [Log    ] [MPMain   ] [INFO ] - Main: Resuming operation
[2014-06-11 15:00:39,913] [Log    ] [MPMain   ] [INFO ] - Main: OnResumeAutomatic - waiting on resume 10 secs
[2014-06-11 15:00:49,913] [Log    ] [MPMain   ] [DEBUG] - Main: OnResumeAutomatic - reopen Database

I'm still pointing towards either (or both) the MPSync & FRITZ!BOX Manager plugins.

It's also good to test this with a longer standby, since i've had issues that were only reproducable after (at least) 1 or 2 hours standby time.
It would be a good test to run your normal setup, and only disable the MPSync & Fritz!Box Manager plugin. Then send the system into standby (before going to bed) and see how it resumes tomorrow.

 

[Requiem]

Okay, I'll be doing so. New logs tomorrow, then.

 

[HomeY]

Okay, I'll be doing so. New logs tomorrow, then.

How was resume this morning?

 

[Requiem]

Unfortunately no resume, MP crashed yestzerday evening, I think there was a database problem again (I didnt rstart after that, had visitors and it was really late...). I have currently rebuilt all databases from scratch and will switch off MPSync. So, tomorrow new logs.

@m3rcury have you heard of others with database corruptions due to MPSync (if this is the cause)?

 

[Requiem]

Okay, forget to start the watchdog (argh) but exported all logs this morning. We are getting close!

What was changed?
1. Rebuilt all databases
2. Undo all changes to the configuration made before, so back to my normal configuration
3. Disabled MPSync and FB-Manager

Started MePo yesterday afternoon, sent it to standby immediately. Resumed yesterday evening about 21:55, resumed ok. Watched Live TV for a while and set MePo back to standby about 00:15 (from live TV). Resumed ok this morning

So, I'll have to slap the plugin devs witg a large trout, I guess :-P