Posts by Stephan Goll


log in
1) Questions and Answers : Unix/Linux : Astropulse errors ... (Message 1138790)
Posted 1076 days ago by Profile Stephan Goll
Well, I've seen http://setiathome.berkeley.edu/forum_thread.php?id=63564 .The sad thing is, that my crunchers show the same problem ... but the "funny" thing is that there is no screensaver. Not even a trace of ... because my crunchers are running headless. It is about http://setiathome.berkeley.edu/hosts_user.php?userid=32167.
Astropulse normally errors out in the first second, but on the atom with CUDA it sometimes crunches a long time until the workunit will finish. Memory and cooling is fine, both boxes are crunching on einstein@home too and the SETI@home Enhanced v5.28 workunits are fine too. http://setiathome.berkeley.edu/workunit.php?wuid=795057091 took a very long time but ended successfully.

My idea is that there may be something wrong in the application (or boinc code?) ... because:

<core_client_version>6.10.58</core_client_version>
<![CDATA[
<message>
process exited with code 193 (0xc1, -63)
</message>
<stderr_txt>
In ap_gfx_main.cpp: in ap_graphics_init(): Starting client.
SIGSEGV: segmentation violation
Stack trace (9 frames):
[0x440158]
[0x7f6484be6f60]
[0x5959e1]
[0x59130e]
[0x59c266]
[0x599ef6]
[0x43d197]
[0x7f6484bde8ba]
[0x7f648494602d]

Exiting...

</stderr_txt>
]]>

and

<core_client_version>6.10.58</core_client_version>
<![CDATA[
<message>
process got signal 11
</message>
<stderr_txt>
In ap_gfx_main.cpp: in ap_graphics_init(): Starting client.

</stderr_txt>
]]>

Again: there is no X running on these crunchers ... and no screensaver because they are running headless.
Is there a way to disable the function call to ap_graphics_init() or make it more tolerant? For me this looks like the main cause of the problem.
Stephan


Copyright © 2014 University of California