100% CPU in 5.0.2

[kltye]

Hi guys,

I recently upgraded to 5.0.1 and then 5.0.2 today, and I've been running this in service mode. Now, whenever a recording finishes, SageTVService (or SageTV.exe if it's running on the console) will take 100% CPU, and it won't record anything else after that. Because I'm now running this as a headless machine, I either remote desktop or VNC into it to control the machine. Prior to this, version 4.1.x didn't exhibit this behavior. I'm running nielm's dynamic menu import alongside the default Sage V5 STV. This only happens when it finishes a recording, schedule or manual.

It's getting pretty annoying because it won't reliably record. Any suggestions on where to look? I'd also be more than happy to provide any additional info needed to help me troubleshoot this.

Thanks!

[ke6guj]

Have you increased the scheduling lookahead from the default 3 days, or have a large (200+) number of favorites?

If either is true, then you may be having problems with the scheduler not being able to compile a schedule fast enough. If it has problems, then it may go into a 100% loop. Either reduce the lookahead or the number of favorites.

[kltye]

Originally Posted by ke6guj

Have you increased the scheduling lookahead from the default 3 days, or have a large (200+) number of favorites?

If either is true, then you may be having problems with the scheduler not being able to compile a schedule fast enough. If it has problems, then it may go into a 100% loop. Either reduce the lookahead or the number of favorites.

nielm's properties in the .properties file says 72 hours to lookahead. There's a scheduling_lookahead property that's set to 432000000. Is that normal?

I think I have only 20 or so favorites; college cable stinks

Thanks.

[ke6guj]

the 72 hours in nielm's properties is to estimate the disk usage for the disk space bar in nielm's dynamic menus, I think. No big deal there.

The scheduling_lookahead at 432000000 is 5 days in milliseconds. Default is 259200000 which is 3 days. I wouldn't think that 20 favorites on a 5 day schedule would cause issues with the scheduler, maybe if you had a really old and slow server. If you wanted to try, shut down sage and set the scheduling_lookahead down to 3, or even 2 days and see if the 100% loop goes away.

[Opus4]

For the 100% cpu issue, you could enable debug logging and submit the results to SageTV, as described in that link.

Originally Posted by ke6guj

If either is true, then you may be having problems with the scheduler not being able to compile a schedule fast enough. If it has problems, then it may go into a 100% loop.

As far as I know, this should not be happening. SageTV still recommends sticking with the default 3 day lookahead (259200000 ms), but I believe the scheduler shouldn't be causing any 100% cou usage issues any more.

- Andy

[kltye]

Thanks for your input guys. Looks like it's going to be a long debug session here; have to see if it spikes to 100% if I try to stop a recording manually...

EDIT: Well it looks like I had enabled debug logging from a previous bug issue. It looks like "Ministry" is taking the 100% CPU; every 3 minutes I get an entry that says "Ministry is waiting for 180 sec". Every once in a while, "Carny is waiting for awhile..." is also entered into the log. I'll be submitting the log to support, but I thought I'd let you guys know...

Thanks again!