PrimeGrid
Please visit donation page to help the project cover running costs for this month

Toggle Menu

Join PrimeGrid

Returning Participants

Community

Leader Boards

Results

Other

drummers-lowrise
1) Message boards : Problems and Help : GPU Task fail with "boinc_get_opencl_ids() failed" (Message 160479)
Posted 111 days ago by Profile Corla99 [Lombardia]
Hi, in the past few days I've noticed that many wus fail on a pc.

I'm using the official docker on a pc with Ubuntu 22.04.1

The stderr output of one wu failed is:
<core_client_version>7.20.2</core_client_version> <![CDATA[ <message> process exited with code 1 (0x1, -255)</message> <stderr_txt> geneferg version 22.12.2 (linux x64, gcc-7.5.0, boinc-7.20.2) Copyright (c) 2022, Yves Gallot genefer is free source code, under the MIT license. Command line: '-boinc -c -n 17 -b 145456638 -f gproof' Error: error: boinc_get_opencl_ids() failed err = -1. 12:55:16 (638171): called boinc_finish(1) </stderr_txt> ]]>


In the logs of the host pc, I've sometimes this message:
audit: type=1400 audit(1676206466.925:165102): apparmor="DENIED" operation="ptrace" profile="docker-default" pid=608146 comm="boinc" requested_mask="read" denied_mask="read" peer="unconfined"

and
AVC apparmor="DENIED" operation="ptrace" profile="docker-default" pid=608146 comm="boinc" requested_mask="read" denied_mask="read" peer="unconfined"


After re-creating the docker container it seems to work...but only for one day or so...

The pc is this
2) Message boards : Problems and Help : Problem with multi-threading and app_config (Message 130207)
Posted 1458 days ago by Profile Corla99 [Lombardia]
Thanks for your help.
At the moment 4 wu are in progress, each with 9 dedicated cores. The cpu usage is around 90%

I tried 2 wu with 18 cores each, but the usage was between 65-75%


Is there a way to know how much cache is in use?
3) Message boards : Problems and Help : Problem with multi-threading and app_config (Message 130035)
Posted 1464 days ago by Profile Corla99 [Lombardia]
Hi, today I moved my pc with 2xE5-2699 on this project. Before doing this, I've disabled the hyper-threading and created an app_config:

<app_config> <app> <name>llr321</name> <fraction_done_exact/> </app> <app_version> <app_name>llr321</app_name> <cmdline>-t 36</cmdline> <avg_ncpus>36</avg_ncpus> </app_version> </app_config>


The wu uses all 36 core, but on htop the usage is almost 45% on each core.

Does anyone know what the problem is?
[Return to PrimeGrid main page]
DNS Powered by DNSEXIT.COM
Copyright © 2005 - 2023 Rytis Slatkevičius (contact) and PrimeGrid community. Server load 2.44, 2.34, 2.47
Generated 4 Jun 2023 | 1:02:44 UTC