Crashes after approx. 1.5 minutes of recording

Bug #995683 reported by David Robertson
54
This bug affects 10 people
Affects Status Importance Assigned to Milestone
Kazam Screencaster
Won't Fix
Undecided
Unassigned

Bug Description

I've installed kazam on my ubuntu 12.04 computer, and it works fine for recording less than about a minute and a half.
After hitting the record button, it records for a little while, then the tray icon disappears.

I ran kazam through the terminal to see what was going on:

➜ ~ kazam
/usr/lib/python2.7/dist-packages/gobject/constants.py:24: Warning: g_boxed_type_register_static: assertion `g_type_from_name (name) == 0' failed
  import gobject._gobject
INFO Indicator - Recording started.
mmap() failed: Cannot allocate memory
Assertion 'b' failed at pulsecore/memblock.c:454, function pa_memblock_acquire(). Aborting.
[1] 11506 abort (core dumped) kazam

I though that perhaps I had ran out of memory, but trying again while watching the system monitor revealed that throughout the whole thing I had plenty of memory left.

Revision history for this message
David Robertson (davidjrobertson) wrote :

Forgot to mention, running `kazam -v` to get the version info outputs this:
kazam 1.0.6 'NCC-2893'

Revision history for this message
Steve Jackson (aearenda) wrote :

It ran for about 16 minutes for me, then failed the same way. However, kazam -v on my system says "kazam 1.3.0-0ubuntu1 'NCC-74656'". I'm using 12.04 as well.

Revision history for this message
Darkn00b (darkn00b1971) wrote :

I am having this exact issue as well with the same error messages. kazam -v reports: kazam 1.0.6 'NCC-2893'

Output of "kazam --debug" attached.

Revision history for this message
Darkn00b (darkn00b1971) wrote :

Sorry. Should have mentioned I am running Linux Mint 12.

Revision history for this message
Adam Bradt (bradticus) wrote :

Have the exact same issue as and I am running 11.10 and using Kazam 1.0.6.

Revision history for this message
David Klasinc (bigwhale) wrote :

Thanks for all the input. Darkn00b, thanks for the debug output. I'd also like debug output from everyone else. Please run kazam --debug in the console and then attach the output from the beginning to the end.

Also, I'd like to know if this happens regardless of the number of sound sources being selected.

Revision history for this message
Adam Bradt (bradticus) wrote :

Just tried again on 1.3 to see if anything was different. Other than the program looking beautiful it only recorded about 5 minutes.

Revision history for this message
Adam Bradt (bradticus) wrote :

Thanks for all your hardwork David. Your attention to issues and feedback is truly something worth recognition. Once this issue is resolved I'll be able to do everything I've ever wanted to do with a screencasting system on Ubuntu. You have a "customer" for life. (Once I have money to work with you will certainly get some donations :D)

Revision history for this message
David Robertson (davidjrobertson) wrote :

I tried a few different settings, see the attachment for the output of kazam --debug for each of the attempts.

Revision history for this message
Darkn00b (darkn00b1971) wrote :

You're welcome David and yes, I have the same issue regardless of number of audio input sources.

Revision history for this message
Steve Jackson (aearenda) wrote :

I have not been able to get a crash with no audio at all, but attached is the debug output after a crash with one audio input.

Revision history for this message
Adam Schwitz (exstntlstfrtn) wrote :

I am also having this problem. I am running ubuntu 12.04. I noticed when I was debugging it at 480x620 it didn't seem to crash, but I didn't wait the full 16 minutes that Steve did. When I changed the resolution to 720x1280 (what I need for my project) it seemed to crash in about 1.5 minutes. Hopefully there is a fix out as I find this to be one of the better interfaces for a screen recorder. Thanks in advance.

Revision history for this message
NofallYaqoo (nofallyaqoo) wrote :

same exactly my problem with kazam (i am using 11.10LTS , and kazam version is 1.3.0)
and i think problem it coused by too much CPU that's why it crash but i am not sure

Revision history for this message
NofallYaqoo (nofallyaqoo) wrote :

ubuntu 11.10LTS *

Revision history for this message
Luis Alvarado (luisalvarado) wrote :

In my case I am using Ubuntu 12.04.1 and Kazam says kazam 1.0.7 'NCC-2893'

It crashes

If I use kazam -d , the debug mode does not detect the sound card, only "Unknown".

Clicking on the Record buttons throws the following in terminal:

/usr/lib/python2.7/dist-packages/gobject/constants.py:24: Warning: g_boxed_type_register_static: assertion `g_type_from_name (name) == 0' failed
  import gobject._gobject
/usr/local/lib/python2.7/dist-packages/kazam/backend/gstreamer.py:156: Warning: value "7" of type `guint' is invalid or out of range for property `threads' of type `guint'
  self.videnc.set_property("threads", self.cores)
INFO Indicator - Recording started.

My Hardware is an Intel I7 2600 with 16GB Ram.

It also crashed around a minute and half to 2 minutes.

Revision history for this message
David Klasinc (bigwhale) wrote :

Are there any updates on this? If anyone is still having problems with, I'd like to ask if they can try to pinpoint if this is happening with VP8 or H264 codec or both?

Revision history for this message
Darkn00b (darkn00b1971) wrote :

David seems to be working on it. There was a code commit for version bump on 11-01-2012. I think the problem here is a lack of people to work on Kazam. David is the only contributor listed on the overview page, so I can't complain about the lack of updates. I know nothing of Python programming or I would volunteer to help. Maybe I can get the word out. Keep up the good work David. For a one man team, you're doing ok.

Revision history for this message
Darkn00b (darkn00b1971) wrote :

Oops. Can someone delete that last comment? I seem to be a derp.

Revision history for this message
David Klasinc (bigwhale) wrote :

Darkn00b, if you're running 12.10, can you check if this is still an issue?

This is apparently PulseAudio bug, but I am not sure why is it occurring nor I am able to reproduce it. I'll be looking into this and checking with PulseAudio guys.

Revision history for this message
dvl (dvaughnluma) wrote :

Also having this issue (Linux Mint 14 / Kazam 1.0.6); however, mine isn't core dumping (at least it doesn't say it is in debug mode). It simply will not allow you to save the video when "finish recording" is selected. It just locks up.

I've had crackling sound issues with Pulseaudio and my Realtek ALC898 on-board sound (ASUS Maximus V Gene mobo), so I assume that this is related. Looks like there are known issues with drivers for this card:

https://bugs.launchpad.net/ubuntu/+source/alsa-driver/+bug/1057095

As noted by some other users, the issue is intermittent, but I've been able to work around it by waiting for the first Kazam process to crash, and then opening a second Kazam process. When both processes are running, I seem to be able to capture video and sound without an issue. I did run a few test passes and was able to capture 10 recordings in sequence without it crashing (the final video was over 20 minutes long).

Looking forward to a fix for this - Thanks for all the hard work you put into this product :)

Revision history for this message
LA2 (lars-aronsson) wrote :

I was able to record 2 minutes of video, but during a 4 minute recording session I suddenly got the following debug output and when I clicked "Finish", nothing was saved. I'm using kazam 1.0.6 from the standard Ubuntu 12.10 on a Lenovo Thinkpad x220i with an Intel Core i3 CPU and 4 GB RAM + 9 GB swap. It has a 1366×768 screen and PulseAudio.

DEBUG PulseAudio - PA - get_source_info_by_index() called. IDX: 1
DEBUG PulseAudio - PA - pa_sourceinfo_cb()
DEBUG PulseAudio - - IDX: 1
DEBUG PulseAudio - - Name: alsa_input.pci-0000_00_1b.0.analog-stereo
DEBUG PulseAudio - - Desc: Built-in Audio Analog Stereo
DEBUG PulseAudio - PA - pa_sourceinfo_cb() -- finished
DEBUG PulseAudio - PA - pa_sourceinfo_cb() -- Hit EOL
DEBUG PulseAudio - PA - pa_sourceinfo_cb() -- EOL no data!
DEBUG PulseAudio - PA - pa_sourceinfo_cb() -- finished
DEBUG Main - Restoring state - volume: A_1 (60.0), A_2 (60.0)
DEBUG Main - Volume 1 changed, new value: 60.0
DEBUG Main - Volume 2 changed, new value: 60.0
DEBUG Main - Codec changed.
DEBUG Main - Cursor capturing ON.
DEBUG Main - Record clicked, invoking Screencast.
/usr/lib/python2.7/dist-packages/gobject/constants.py:24: Warning: g_boxed_type_register_static: assertion `g_type_from_name (name) == 0' failed
  import gobject._gobject
DEBUG GStreamer - Capture Cursor: True
DEBUG GStreamer - Framerate : True
DEBUG GStreamer - Coordinates: 0 0 1365 767
DEBUG GStreamer - Codec: VP8/WEBM
DEBUG GStreamer - Audio1 Source:
  alsa_input.pci-0000_00_1b.0.analog-stereo
DEBUG GStreamer - Filesink: /tmp/kazam_OhzVRg.movie
DEBUG GStreamer - Pipline - Video + Audio
DEBUG Main - Hiding main window.
DEBUG Main - Start request.
INFO Indicator - Recording started.
DEBUG GStreamer - Generating dot file.
DEBUG GStreamer - Setting STATE_PLAYING
mmap() failed: Cannot allocate memory
mmap() failed: Cannot allocate memory
(repeated 128 times)
mmap() failed: Cannot allocate memory
DEBUG GStreamer - Received an error message.
DEBUG GStreamer - Received an error message.
(repeated 21368 times)
DEBUG GStreamer - Received an error message.

***MEMORY-ERROR***: kazam[5706]: GSlice: failed to allocate 504 bytes (alignment: 512): Cannot allocate memory

DEBUG GStreamer - Received an error message.

Revision history for this message
LA2 (lars-aronsson) wrote :

I was wrong, the failed recording session was 8 minutes. I was able to rescue the video from the /tmp/ file mentioned in the debug output. I attach the /tmp/kazam_debug.dot file which seems to have been created when the program started.

Revision history for this message
Alex Wochna (alex-wochna) wrote :
Download full text (9.1 KiB)

awochna@ubuntu:~$ kazam -d
DEBUG Kazam - Starting ...
DEBUG Main - Setting variables.
DEBUG PulseAudio - PA - Starting mainloop.
DEBUG PulseAudio - PA - Getting API.
DEBUG PulseAudio - PA - Setting context.
DEBUG PulseAudio - PA - Set state callback.
DEBUG PulseAudio - PA - Connecting to server.
DEBUG PulseAudio - PA - Start mainloop.
DEBUG PulseAudio - PA - Connected.
DEBUG Main - Connecting indicator signals.
DEBUG Main - Main Window UI setup.
DEBUG Main - Getting Audio sources.
DEBUG PulseAudio - PA - get_audio_sources() called.
DEBUG PulseAudio - PA - pa_sourcelist_cb()
DEBUG PulseAudio - - IDX: 0
DEBUG PulseAudio - - Name: alsa_output.pci-0000_01_00.1.hdmi-stereo.monitor
DEBUG PulseAudio - - Desc: Monitor of RV710/730 HDMI Audio [Radeon HD 4000 series] Digital Stereo (HDMI)
DEBUG PulseAudio - PA - pa_sourcelist_cb()
DEBUG PulseAudio - - IDX: 1
DEBUG PulseAudio - - Name: alsa_output.pci-0000_00_1b.0.analog-stereo.monitor
DEBUG PulseAudio - - Desc: Monitor of Built-in Audio Analog Stereo
DEBUG PulseAudio - PA - pa_sourcelist_cb()
DEBUG PulseAudio - - IDX: 2
DEBUG PulseAudio - - Name: alsa_input.pci-0000_00_1b.0.analog-stereo
DEBUG PulseAudio - - Desc: Built-in Audio Analog Stereo
DEBUG PulseAudio - PA - pa_sourcelist_cb()
DEBUG PulseAudio - - IDX: 3
DEBUG PulseAudio - - Name: alsa_input.usb-046d_081b_BE503210-02-U0x46d0x81b.analog-mono
DEBUG PulseAudio - - Desc: Webcam C310 Analog Mono
DEBUG PulseAudio - PA - pa_sourcelist_cb() -- finished
DEBUG Main - Device found:
DEBUG Main - - 0
DEBUG Main - - alsa_output.pci-0000_01_00.1.hdmi-stereo.monitor
DEBUG Main - - Monitor of RV710/730 HDMI Audio [Radeon HD 4000 series] Digital Stereo (HDMI)
DEBUG Main - Device found:
DEBUG Main - - 1
DEBUG Main - - alsa_output.pci-0000_00_1b.0.analog-stereo.monitor
DEBUG Main - - Monitor of Built-in Audio Analog Stereo
DEBUG Main - Device found:
DEBUG Main - - 2
DEBUG Main - - alsa_input.pci-0000_00_1b.0.analog-stereo
DEBUG Main - - Built-in Audio Analog Stereo
DEBUG Main - Device found:
DEBUG Main - - 3
DEBUG Main - - alsa_input.usb-046d_081b_BE503210-02-U0x46d0x81b.analog-mono
DEBUG Main - - Webcam C310 Analog Mono
DEBUG Main - Getting Video sources.
DEBUG Main - Found 1 monitors.
DEBUG Main - Monitor 0 - X: 0, Y: 0, W: 1366, H: 768
DEBUG Main - Video Toggled.
DEBUG Main - Audio1 Toggled.
DEBUG Main - Restoring state - sources: V (0), A_1 (-1), A_2 (1)
DEBUG Main - Video changed.
DEBUG Main - New Video: {'y': 0, 'x': 0, 'height': 768, 'width': 1366}
DEBUG Main - Audio Changed.
DEBUG Main - - A_1 -1
DEBUG Main - - A_2 1
DEBUG Main - - PA Audio1 IDX: 3
DEBUG Main - - PA Audio2 IDX: 1
DEBUG PulseAudio - PA - get_source_info_by_index() called. IDX: 3
DEBUG PulseAudio - PA - pa_sourceinfo_cb()
DEBUG PulseAudio - - IDX: 3
DEBUG PulseAudio - - Name: alsa_input.usb-046d_081b_BE503210-02-U0x46d0x81b.analog-mono
DEBUG PulseAudio - - Desc: Webcam C310 Analog Mono
DEBUG PulseAudio - PA - pa_sourceinf...

Read more...

Revision history for this message
Alex Wochna (alex-wochna) wrote :

Also forgot to mention, I'm recording what I'm doing in a Google Chrome tab and after the lock up and auto-killing the process, Chrome back back with the "He's Dead Jim!" page (the page failed to load because of lack of memory or was terminated for another reason)

Revision history for this message
Tom Andrews (gta4lj) wrote :

I get the same problem, when i record, after a random amount of time, the tray icon disappears and there is no file left in the target folder, i use mint 14 cinnamon , if anyone could tell me where the debug output is placed( if it is atall ? ) then ill upload it. I am currently using the unstable ppa version , don't know off the top of my head. I know my way round a terminal but not specific things.

Thanks- Tom

Revision history for this message
David Klasinc (bigwhale) wrote :

Alex and Tom, this bug was opened for Kazam 1.0.x which is mostly obsolete and a lot of stuff was rewritten. Please try the latest stable version from the stable PPA if the problem is still there. If it is, please open one more bug report and attach debug output from kazam when you run it in terminal with --debug option.

Changed in kazam:
status: New → Won't Fix
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Duplicates of this bug

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.