<!--#include virtual="header.txt"-->

<h1>Intel Knights Landing (KNL) User and Administrator Guide</h1>

<h2 id="overview">Overview<a class="slurm_link" href="#overview"></a></h2>

<p>This document describes the unique features of Slurm on the computers with
the Intel Knights Landing processor.
You should be familiar with the Slurm's mode of operation on Linux clusters
before studying the relatively few differences in Intel KNL system operation
described in this document.</p>

<h2 id="user_tools">User Tools
<a class="slurm_link" href="#user_tools"></a>
</h2>

<p>The desired NUMA and MCDRAM modes for a KNL processor should be specified
using the -C or --constraint option of Slurm's job submission commands: salloc,
sbatch, and srun. Currently available NUMA and MCDRAM modes are shown in the
table below. Each node's available and current NUMA and MCDRAM modes are
visible in the "available features" and "active features" fields respectively,
which may be seen using the scontrol, sinfo, or sview commands.
Note that a node may need to be rebooted to get the desired NUMA and MCDRAM
modes and nodes may only be rebooted when they contain no running jobs
(i.e. sufficient resources may be available to run a pending job, but until
the node is idle and can be rebooted, the pending job may not be allocated
resources). Also note that the job will be charged for resources from the time
of resource allocation, which may include time to reboot a node into the
desired NUMA and MCDRAM configuration.</p>

<p>Slurm supports a very rich set of options for the node constraint options
(exclusive OR, node counts for each constraint, etc.).
See the man pages for the salloc, sbatch and srun commands for more information
about the constraint syntax.
Jobs may specify their desired NUMA and/or MCDRAM configuration. If no
NUMA and/or MCDRAM configuration is specified, then a node with any possible
value for that configuration will be used.</p>

<table width="100%" border=1 cellspacing=0 cellpadding=4>
<tr>
  <th width="15%">Type</th>
  <th width="15%">Name</th>
  <th width="70%">Description</th>
</tr>
<tr><td>MCDRAM</td><td>cache</td><td>All of MCDRAM to be used as cache</td></tr>
<tr><td>MCDRAM</td><td>equal</td><td>MCDRAM to be used partly as cache and partly combined with primary memory</td></tr>
<tr><td>MCDRAM</td><td>flat</td><td>MCDRAM to be combined with primary memory into a "flat" memory space</td></tr>
<tr><td>NUMA</td><td>a2a</td><td>All to all</td></tr>
<tr><td>NUMA</td><td>hemi</td><td>Hemisphere</td></tr>
<tr><td>NUMA</td><td>snc2</td><td>Sub-NUMA cluster 2</td></tr>
<tr><td>NUMA</td><td>snc4</td><td>Sub-NUMA cluster 4 (<a href="#note">NOTE</a>)</td></tr>
<tr><td>NUMA</td><td>quad</td><td>Quadrant (<a href="#note2">NOTE</a>)</td></tr>
</table>

<p>Jobs requiring some or all of the KNL high bandwidth memory (HBM) should
explicitly request that memory using Slurm's Generic RESource (GRES) options.
The HBM will always be known by Slurm GRES name of "hbm".
Examples below demonstrate use of HBM.</p>

<p>Sorting of the free cache pages at step startup using Intel's zonesort
module can be configured as the default for all steps using the
"LaunchParameters=mem_sort" option in the slurm.conf file.
Individual job steps can enable or disable sorting using the "--mem-bind=sort"
or "--mem-bind=nosort" command line options for srun.
Sorting will be performed only for the NUMA nodes allocated to the job step.</p>

<p><a id="note"><b>NOTE</b></a>: Slurm provides limited support
for restricting use of HBM. At some point in the future, the amount of HBM
requested by the job will be compared with the total amount of HBM and number of
memory-containing NUMA nodes available on the KNL processor. The job will then
be bound to specific NUMA nodes in order to limit the total amount of HBM
available to the job, and thus reserve the remaining HBM for other jobs running
on that KNL processor.</p>

<p><a id="note2"><b>NOTE</b></a>: Slurm can only
support homogeneous nodes (e.g. the same number of cores per NUMA node).
KNL processors with <u>68 cores</u> (a subset of KNL models) will not have
homogeneous NUMA nodes in snc4 mode, but each NUMA node will have
either 16 or 18 cores. This will result in Slurm using the lower core count,
finding a total of 256 threads rather than 272 threads and setting the node
to a DOWN state.</p>

<h3 id="accounting">Accounting<a class="slurm_link" href="#accounting"></a></h3>

<p>If a node requires rebooting for a job's required configuration, the job
will be charged for the resource allocation from the time of allocation through
the lifetime of the job, including the time consumed for booting the nodes.
The job's time limit will be calculated from the time that all nodes are ready
for use.
For example, a job with a 10 minute time limit may be allocated resources at
10:00:00.
If the nodes require rebooting, they might not be available for use until
10:20:00, 20 minutes after allocation, and the job will begin execution at
that time.
The job must complete no later than 10:30:00 in order to satisfy its time limit
(10 minutes after execution actually begins).
However, the job will be charged for 30 minutes of resource use, which includes
the boot time.</p>

<h3 id="use_case">Sample Use Cases
<a class="slurm_link" href="#use_case"></a>
</h3>

<pre>
$ sbatch -C flat,a2a -N2 --gres=hbm:8g --exclusive my.script
$ srun --constraint=hemi,cache -n36 a.out
$ srun --constraint=flat --gres=hbm:2g -n36 a.out

$ sinfo -o "%30N %20b %f"
NODELIST       ACTIVE_FEATURES  AVAIL_FEATURES
nid000[10-11]
nid000[12-35]  flat,a2a         flat,a2a,snc2,hemi
nid000[36-43]  cache,a2a        flat,equal,cache,a2a,hemi
</pre>

<h3 id="topology">Network Topology
<a class="slurm_link" href="#topology"></a>
</h3>

<p>Slurm will optimize performance using those resources available without
rebooting. If node rebooting is required, then it will optimize layout with
respect to network bandwidth using both nodes currently in the desired
configuration and those which can be made available after rebooting.
This can result in more nodes being rebooted than strictly needed, but will
improve application performance.</p>

<p>Users can specify they want all resources allocated on a specific count of
leaf switches (Dragonfly group) using Slurm's <b>--switches</b> option.
They can also specify how much additional time they are willing to wait for
such a configuration. If the desired configuration can not be made available
within the specified time interval, the job will be allocated nodes optimized
with respect to network bandwidth to the extent possible. On a Dragonfly
network, this means allocating resources over either single group or
distributed evenly over as many groups as possible. For example:</p>
<pre>
srun --switches=1@10:00 N16 a.out
</pre>
<p>Note that system administrators can disable use of the <b>--switches</b>
option or limit the amount of time the job can be deferred using the
<b>SchedulerParameters</b> <b>max-switch-wait</b> option.</p>

<h3 id="boot_problems">Booting Problems
<a class="slurm_link" href="#boot_problems"></a>
</h3>

<p>If node boots fail, those nodes are drained and the job is requeued so that
it can be allocated a different set of nodes. The nodes originally allocated
to the job will remain available to the job, so likely a small number of
additional nodes will be required.</p>

<h2 id="administration">System Administration
<a class="slurm_link" href="#administration"></a>
</h2>

<p>Four important components are required to use Slurm on an Intel KNL system.</p>
<ol>
<li>Slurm needs a mechanism to determine the node's current topology (e.g.
how many NUMA exist and which cores are associated with each NUMA). Slurm
relies upon <a href="http://www.open-mpi.org/projects/hwloc/">
Portable Hardware Locality (HWLOC)</a> for this functionality. Please install
HWLOC before building Slurm.</li>

<li>The node features plugin manages the available and active features
information available for each KNL node.</li>

<li>A configuration file is used to define various timeouts, default
configuration, etc. The configuration file name and contents will depend upon
the node features plugins used. See the <a href="knl.conf.html">knl.conf</a>
man page for more information.</li>

<li>A mechanism is required to boot nodes in the desired configuration. This
mechanism must be integrated with existing Slurm infrastructure for
<a href="sbatch.html">rebooting nodes on user request (--reboot)</a>.</li>
</ol>

<p>In addition, there is a DebugFlags option of "NodeFeatures" which will
generate detailed information about KNL operations.</p>

<p>The KNL-specific available and active features are configured differently
based upon the plugin configured.<br>
<u>For the knl_generic plugin</u>, KNL-specific features should be defined
in the "slurm.conf" configuration file. When the slurmd daemon starts on each
compute node, it will update the available and active features as needed.<br>
Features which are not KNL-specific (e.g. rack number, "knl", etc.) will be
copied from the node's "Features" configuration in "slurm.conf" to both the
available and active feature fields and not modified by the NodeFeatures
plugin.</p>

<p><b>NOTE</b>: For Dell KNL systems you must also include the
<i>SystemType=Dell</i> option for successful operation and will likely need to
increase the <i>SyscfgTimeout</i> to allow enough time for the command to
successfully complete.  Experience at one site has shown that a 10 second
timeout may be necessary, configured as <i>SyscfgTimeout=10000</i>.</p>

<p>Slurm does not support the concept of multiple NUMA nodes
within a single socket. If a KNL node is booted with multiple NUMA, then each
NUMA node will appear in Slurm as a separate socket.
In the slurm.conf configuration file, set node socket and
core counts to values which are appropriate for some NUMA mode to be used on the
node. When the node boots and the slurmd daemon on the node starts, it will
report to the slurmctld daemon the node's actual socket (NUMA) and core counts,
which will update Slurm data structures for the node to the values which are
currently configured.
Note that Slurm currently does not support the concept of
differing numbers of cores in each socket (or NUMA node). We are currently
working to address these issues.</p>

<h3 id="operation">Mode of Operation
<a class="slurm_link" href="#operation"></a>
</h3>

<ol>
<li>The node's configured "Features" are copied to the available and active
feature fields.</li>
<li>The node features plugin determines the node's current MCDRAM and NUMA
values as well as those which are available and adds those values to the node's
active and available feature fields respectively. Note that these values may
not be available until the node has booted and the slurmd daemon on the
compute node sends that information to the slurmctld daemon.</li>
<li>Jobs will be allocated nodes already in the requested MCDRAM and NUMA mode
if possible. If insufficient resources are available with the requested
configuration then other nodes will be selected and booted into the desired
configuration once no other jobs are active on the node. Until a node is idle,
its configuration can not be changed. Note that node reboot time is roughly
on the order of 20 minutes.</li>
</ol>

<h3 id="config">Generic Cluster Configuration
<a class="slurm_link" href="#config"></a>
</h3>

<p>All other clusters should have NodeFeaturesPlugins configured to "knl_generic".
This plugin performs all operations directly on the compute nodes using Intel's
<i>syscfg</i> program to get and modify the node's MCDRAM and NUMA mode and
uses the Linux <i>reboot</i> program to reboot the compute node in order for
modifications in MCDRAM and/or NUMA mode to take effect.
Make sure that RebootProgram is defined in the slurm.conf file.
This plugin currently does <u>not</u> permit the specification of ResumeProgram,
SuspendProgram, SuspendTime, etc. in slurm.conf, however that limitation may
be removed in the future (the ResumeProgram currently has no means of changing
the node's MCDRAM and/or NUMA mode prior to reboot).</p>

<p><b>NOTE</b>: The syscfg program reports the MCDRAM and NUMA mode to be used
when the node is next booted. If the syscfg program is used to modify the MCDRAM
or NUMA mode of a node, but it is not rebooted, then Slurm will be making
scheduling decisions based upon incorrect state information. If you want to
change node state information outside of Slurm then use the following procedure:
<ol>
<li>Drain the nodes of interest</li>
<li>Change their MCDRAM and/or NUMA mode</li>
<li>Reboot the nodes, then</li>
<li>Restore them to service in Slurm</li>
</ol>
</p>

<h4>Sample knl_generic.conf File</h4>

<pre>
# Sample knl_generic.conf
SyscfgPath=/usr/bin/syscfg
DefaultNUMA=a2a         # NUMA=all2all
AllowNUMA=a2a,snc2,hemi
DefaultMCDRAM=cache     # MCDRAM=cache
</pre>

<h4>Sample slurm.conf File</h4>

<pre>
# Sample slurm.conf
NodeFeaturesPlugins=knl_generic
DebugFlags=NodeFeatures
GresTypes=hbm
RebootProgram=/sbin/reboot
...
Nodename=default Sockets=1 CoresPerSocket=68 ThreadsPerCore=4 RealMemory=128000 Feature=knl
NodeName=nid[00000-00127] State=UNKNOWN
</pre>


<p style="text-align:center;">Last modified 13 March 2024</p>

<!--#include virtual="footer.txt"-->