Occasionally, for "Period Search Application v102.12 (arm_android_pie) arm-android-linux-gnu", the task will error out in under 3 seconds, and the stderr will report:

<core_client_version>7.24.1</core_client_version>
<![CDATA[
<message>
process exited with code 2 (0x2, -254)</message>
<stderr_txt>
stackdumps unavailable

Error: Number of lc points is greater than POINTS_MAX = 1000

</stderr_txt>
]]>

---

There used to be a limit of 1000 light points in each work unit. Apps for some OSs got upgraded to handle more (Windows, Unix). Other apps have not yet been upgraded. I know macOS has the same issue.

I confirm on mac OS :

Stderr output

<core_client_version>7.24.1</core_client_version>
<![CDATA[
<message>
process exited with code 2 (0x2, -254)</message>
<stderr_txt>

Error: Number of lc points is greater than POINTS_MAX = 1000

</stderr_txt>
]]>

I (re)started to crunch this afternoon (CPU for mac), 32 tasks in success and 17 with this error. On task detail it says "0 seconds processed" so I hope it's true.

Note : I just realized that the mac app version is from... 2013 !! for windows or linux it's 2020 or 2022... no comment :/

Finally I'm lucky that it is still working somehow !! (on other projects old boinc mac apps ceased working with Apple 32b-->64b move)

Mmmm and since this morning the project refuses to send me tasks

Sam 18 nov 12:58:42 2023 | Asteroids@home | Requesting new tasks for CPU and AMD/ATI GPU
Sam 18 nov 12:58:44 2023 | Asteroids@home | Scheduler request completed: got 0 new tasks
Sam 18 nov 12:58:44 2023 | Asteroids@home | Project has no tasks available

even though server status claims to have plenty to send...

I assume that this same error rate (now 145 success and 105 errors) is the cause and the project automatically restrains tasks for that machine ? (I know some projects have such policy to avoid buggy machine to waste cycles).

In the task detail I have no calculation time on the errored tasks (not 0: no figure) so I don't know if it's true and they fail immediately, or if it actually wastes CPU cycles before generating that error.

Anyway, back to other projects.