Lost frames, 50 sec section length, & zm.ffmpeg.cpp error
Posted: Wed Jun 19, 2019 1:37 pm
I am currently running 1.33.9~20190612100557-xenial
Code: Select all
2019-06-19 07:03:02 zmc_m2 30426 WAR error while decoding MB 11 40, bytestream -26 zm_ffmpeg.cpp 64 2019-06-19 07:02:02 zmc_m5 30225 WAR error while decoding MB 78 37, bytestream -5 zm_ffmpeg.cpp 64 2019-06-19 07:01:02 zmc_m4 29951 WAR error while decoding MB 42 9, bytestream -54 zm_ffmpeg.cpp 64 2019-06-19 07:01:02 zmc_m2 29928 WAR error while decoding MB 67 20, bytestream -21 zm_ffmpeg.cpp 64 2019-06-19 07:00:02 zmc_m2 20434 WAR error while decoding MB 55 16, bytestream -18 zm_ffmpeg.cpp 64 2019-06-19 07:00:02 zmc_m3 20454 WAR error while decoding MB 39 28, bytestream -60 zm_ffmpeg.cpp 64
I have a cron job that changes the runstate using: /usr/bin/zmpkg.pl "New State"
This morning upon the change of state my continuous events were all reduced to approximately 50 seconds length, with a ten second loss between events, and the above error being generated in the log.
I tried to restart the state through the console interface; the problem continued. I rebooted the machine; the problem continued.
Finally, I used the console interface to change to a different state, and the events were created with the section length specified in the settings, events had a slight frame overlap at beginning and end, and the error messages stopped. I then used the console to change back to the state that had been specified by the cron job when things stopped working properly, and everything seems to be working as expected in that state also . . .
I don't understand how the problem could persist through the reapplication of the state, a reboot, and then go away with a change to a different state?
*** On further investigation, I see that after the cron job runs, I am getting this broken behaviour for an hour. However, after this hour, the sections begin and end properly again. What am I doing wrong?