From 49ef3bfc3d943f63c3d3e24d904a6b75d9694544 Mon Sep 17 00:00:00 2001 From: Morris Jette <jette@schedmd.com> Date: Thu, 1 Dec 2016 12:52:27 -0700 Subject: [PATCH] Clarify core_spec selection algorithm --- doc/html/core_spec.shtml | 13 ++++++++++--- 1 file changed, 10 insertions(+), 3 deletions(-) diff --git a/doc/html/core_spec.shtml b/doc/html/core_spec.shtml index 1d81ffdf79e..14c2b2cda60 100644 --- a/doc/html/core_spec.shtml +++ b/doc/html/core_spec.shtml @@ -36,8 +36,15 @@ commands will reflect all CPUS on all allocated nodes as will the job's accounting.</p> <h2>Core Selection</h2> -<p>Cores selected for specialization will always follow the same assignment -algorithm. +<p>The specific resources to be used for specialization may be identified using +the <i>CPUSpecList</i> configuration parameter associated with each node in +the <i>slurm.conf</i> file. +Note that the <i>core_spec/cray</i> does not currently support identification of +specific cores, so that plugin should not be used in conjunction with the +<i>CPUSpecList</i> configuration parameter, even on Cray systems. +If <i>CoreSpecCount</i> is configured, but not <i>CPUSpecList</i>, the cores +selected for specialization will always follow the assignment algorithm +described below. The first core selected will be the highest numbered core on the highest numbered socket. Subsequent cores selected will be the highest numbered core on the lower @@ -86,6 +93,6 @@ If the core count is not configured and the CPUs value is configured to the count of hyperthreads, then hyperthreads rather than cores will be reserved for system use.</p> -<p style="text-align:center;">Last modified 6 June 2014</p> +<p style="text-align:center;">Last modified 1 December 2016</p> <!--#include virtual="footer.txt"--> -- GitLab