TV Server Crashes After Recording Ends (1 Viewer)

[Sigpi007]

Hey @georgius and @mm1352000 ... just wanted to provide an update on my experience with this new filter. It generates some excess heat in my PC!

In the above thread... I did a lot of the testing on a dev environment which worked to get the new filter working in my particular situation (cable card / networked tuner). That worked great to remove the major issues. Now I've tried this filter out in my "production environment"... aka the family room TV... and noticed that the new filter increases the amount of CPU utilization by a reasonably noticeable amount during playback. I was watching a Blackhawks game and stopped to apply the new filter, and then started the game back up and noticed that my heatsink fan kicked on high after about 5 minutes of watching. I checked perfmon and saw that my CPU utilization was around 18-20% for MP during playback, when typically its more like 13% - 16% in my setup using the 1.x filter.

I know its not much, but the difference is enough to make the heat output jump enough to make the fan scream. I'm even more worried about this when recordings are happening in the background. (None were active during the game I was watching.)

Any thoughts?

Logs from the "production" setup attached. I had the new filter running from 21:17 to 21:35 if that helps.

(PS... Merry Christmas / happy holidays to you and yours. Thanks for all your work making MP a really awesome product.)

 

[georgius]

I need to try it on my low powered Atom HTPC, after that we'll see if there is some problem. On my dev Core i5 I don't see any performance problem.

 

[georgius]

Performance problem confirmed on my production Atom D510, cca 5-7% about standard DVB-IP filter. Unfortunately, testing window on my production system is very small, due to almost all time watching or recording And what is worse, I don't have any idea where to start finding performance problem.

 

[Sigpi007]

At least we know its not an isolated case.

Funny though... my dev system is a core i5 as well... and my production system is not an Atom... but not far off from that type of processor since its an AMD A8 APU (with no discrete video card). Both are made for low power, small form factor setups (like very quiet HTPCs.)

 

[georgius]

Good news In my latest highly experimental filter with UDP protocol and Atom processor I have approx. 8-10% lower CPU usage then with stock IPTV filter. So, please be patient until I rework RTSP protocol and after that we can make some tests.

 

[Sebastiii]

That's good in recent commit i wasn't able to tune rtsp so i revert back to some previous commit, i have let you log few week ago (on irc).
And happy new year

 

[georgius]

That's good in recent commit i wasn't able to tune rtsp so i revert back to some previous commit, i have let you log few week ago (on irc).
And happy new year

You have to switch to MP1-4155-FEAT-Merge_of_IPTV_filter_and_MediaPortal_Url_Source_Splitter_2_1_4_5424 branch, in original branch I try to implement MPEG2 TS muxer (postponed until we tweak performance of filter).

 

[georgius]

@Sigpi007 , try attached filter version and let me know, if performance is better or not.

 

[Sigpi007]

Just saw this message. Will test and post results in next 12 hours (when I can get time on that box from the family).

 

[Sigpi007]

try attached filter version and let me know, if performance is better or not.

Performance not good. Had a stutter in playback about once per second, so it wasn't watchable. The CPU didn't seem to spike though. CPU utilization was actually lower than the 1.x filter, but not sure this is a good comparison until the appropriate playback is being observed.

Logs attached from 1.X filter and then new filter posted above.