Sage killed occasionally

[stanger89]

So over the past few months, I've occasionally found in the morning that Sage is no longer running. When I looked in the sagetv_x.log file there was nothing out of the ordinary, the log just ended. There were no hs_err_pid files either.

I just had it happen again this morning but this time I tried checking the unRAID/Docker log and found this:

./startsagecore: line 55:   114 Killed                   java -Djava.awt.headless=$HEADLESS $JAVAMEM  -XX:+UseAdaptiveSizePolicy -XX:MaxGCPauseMillis=25 -XX:GCTimeRatio=24  $(if [[ $EUID -eq 0 ]]; then echo '-XX:ThreadPriorityPolicy=1'; fi)  $JAVAOPTS -cp Sage.jar:JARs/lucene-core-3.6.0.jar:/:$(echo JARs/*.jar |  sed 's/  */:/g') sage.Sage 0 0 x "sagetv Sage.properties" 0>&-

Not really sure what else to check for, thoughts?

[wayner]

I had a similar situation in the past where certain times of image files being displayed by the Screen Saver were causing Java to crash.

Would you have clients like extenders that stay on trying to render the screen saver?

[stanger89]

There shouldn't be any. All my setups have shutdown macros that turn off the extenders along with everything else. That and there was nothing in the log, it ended like this:

Sat 1/6 4:30:30.222 [Seeker@22be474] Seeker waiting for 84 mins.
Sat 1/6 4:31:20.964 [Ministry@756fc8b9] Ministry is waiting for 180 sec
Sat 1/6 4:31:21.047 [ThreadMonitor@6577a6ae] Top 10 Cumulative CPU(sec) AsyncPropSaver:907 ReProcessHook:480 ReProcessHook:398 ReProcessHook:145 Timer-5:89 ReProcessHook:51 EPG:42 AsyncPropSaver:41 Seeker:27 ReProcessHook:24
Sat 1/6 4:31:21.047 [ThreadMonitor@6577a6ae] Top 10 %CPU (last 300s)
Sat 1/6 4:34:29.269 [Ministry@756fc8b9] Ministry is waiting for 180 sec

Looking at the current log, which has an extender running the screensaver right now, I can see each picture loaded in the log. On top of that, since I've been at home on vacation, I've had the screensaver running on one extender basically all day, every day, and Sage has never crashed.

Interestingly, looking at the unraid/Docker log, it seems like it's always between 4 and 5am that Sage dies or is killed....

....

So I checked my syslog, and it looks like Bunker (File Integrity Plugin) happens to be running, and running out of memory at the same time, so I reduced my disk cache settings based on some recommendations here:
https://lime-technology.com/forums/t...lp-please-635/

We'll see what happens.

[wayner]

Ok, good luck. I bet it was that memory issue - I think I have seen that before in the forums - either here or Lime.

[BobPhoenix]

On one of my servers I was constantly getting SageTV Docker crashes. When I disabled CommercialDetector in the docker so that ComSkip didn't get called my problems disappeared. I now use DirMon2 running on a Windows VM calling ComSkip also installed on the VM. My other SageTV Docker instance hasn't had any problems but it is only running comskip on MPG files not H.264 TS files like the one that crashed approximately every week. Also those H.264 TS files are HD-PVR recordings with 5.1 sound. It is possible that if I only recorded Stereo in the H.264 TS files I wouldn't have had problems either but I'm not going to give that up.

[stuckless]

Are you running some sort of backup utility on unRAID... some of those kill the docker containers, then back up, and then optionally restart.

[stanger89]

Originally Posted by stuckless

Are you running some sort of backup utility on unRAID... some of those kill the docker containers, then back up, and then optionally restart.

Nope, well I've got crashplan, but that doesn't restart things. I'm thinking it's FIP/Bunker OOM, I had a couple of those pop up in the log right in the window when Sage would get killed.