Project

General

Profile

Bug #1755

Tvheadend becomes unresponsive

Added by Martin Mrvka over 11 years ago. Updated over 10 years ago.

Status:
Rejected
Priority:
Normal
Assignee:
-
Category:
Timeshift
Target version:
-
Start date:
2013-06-14
Due date:
% Done:

0%

Estimated time:
Found in version:
3.5.64~g5e79502
Affected Versions:

Description

Sometimes, after running for a couple of days, tvheadend becomes unresponsive.

This means that the process itself is still running, in this case not consuming much resources (no cpu hog or huge memory consumption). However there is no response on http or htsp port anymore.

Connection to both ports still work, but sending data (eg. "GET / HTTP/1.0" on the http port) doesn't generate any reponse (expecting a HTTP 401 in this case).

The last time this happened there was the following log message:

[ xbmc | XBMC Media Center ]: failed to serialize data

History

#1

Updated by Andy Brown over 11 years ago

Hi,
Can you check and see if you get anything in your tvheadend logging? Probably to syslog, especially when your XBMC or web client try to connect to see if any messages get logged.

If you restart just the tvheadend daemon does it recover and start working again until another day or so?

We might need it ran to generate more verbose logs, but if you let me know on the above first and see if we can spot whats going on.

#2

Updated by Andy Brown over 11 years ago

  • Status changed from New to Need feedback
#3

Updated by Martin Mrvka over 11 years ago

Hi,

unfortunately there is no more logging afterwards when tvheadend got stuck. I turned debug mode and there is an additional line before the one already posted.

However all logs (even in debug mode) available are:

Jun 30 14:18:06 mediastation tvheadend[16261]: timeshift: ts 150 eob revert to live mode
Jun 30 14:18:06 mediastation tvheadend[16261]: htsp: X.X.X.X [ xbmc | XBMC Media Center ]: failed to serialize data

I configured monit to restart tvheadend in this case, which works fine. So yes, a restart works fine. I need to disable monit in order to do some strace to see if it's a blocking file handle in the critical path or so. Maybe there is a relation to timeshift (which I'm using always on).

#4

Updated by Martin Mrvka over 11 years ago

I switched back to 3.4.27~gfbda802~raring and the problem still exists.

TCP connects to HTSP port and webinterface are possible, but no response. TVH is running with debug log, however the only markable output is

timeshift: ts 2 eob revert to live mode
(the number varies) which comes before every lockup.

#5

Updated by Adam Sutton almost 11 years ago

  • Category changed from General to Timeshift

No idea what's going on here, but I'm going to lump in with timeshift as that's the only relevant logging message in the report.

Adam

#6

Updated by Adam Sutton over 10 years ago

  • Status changed from Need feedback to Rejected

Really not sure how to proceed with this, I'm not seeing it myself (not that I'm a massive timeshift user) and no other reports. Plus lots of fixes have been applied in master that could relate.

Adam

Also available in: Atom PDF