Skip to content
Snippets Groups Projects
Commit 20b0ac7d authored by Morris Jette's avatar Morris Jette
Browse files

Major update to NEWS about new releases

parent ab218ba2
No related branches found
No related tags found
No related merge requests found
Hide whitespace changes
Inline Side-by-side
Showing
with 40 additions and 55 deletions
doc/html/news.shtml
+ 40
− 55
View file @ 20b0ac7d
......@@ -4,74 +4,63 @@
<h2>Index</h2>
<ul>
<li><a href="#26">Slurm Version 2.6, June 2013</a></li>
<li><a href="#1403">Slurm Version 14.03, March 2014</a></li>
<li><a href="#1409">Slurm Version 14.09 and beyond</a></li>
<li><a href="#1411">Slurm Version 14.11, November 2014</a></li>
<li><a href="#1508">Slurm Version 15.08 and beyond</a></li>
</ul>
<h2><a name="26">Major Updates in Slurm Version 2.6</a></h2>
<p>SLURM Version 2.6 was released in July 2013.
<h2><a name="1403">Major Updates in Slurm Version 14.03</a></h2>
<p>SLURM Version 14.03 was released in March 2014.
Major enhancements include:
<ul>
<li>Support for job arrays, which increases performance and ease of use
for sets of similar jobs.</li>
<li><a href="hdf5_profile_user_guide.html">Job profiling</a> capability added
to record a wide variety of job characteristics for each task on a user
configurable periodic basis. Data currently available includes CPU use,
memory use, energy use, Infiniband network use, Lustre file system use,
etc.</li>
<li>Support for MPICH2 using PMI2 communications interface with much greater
scalability.</li>
<li>Prolog and epilog support for advanced reservations.</li>
<li>Much faster throughput for job step execution with --exclusive option. The
srun process is notified when resources become available rather than
periodic polling.</li>
<li>Support improved for Intel MIC (Many Integrated Core) processor.</li>
<li>Advanced reservations with hostname and core counts now supports asymmetric
reservations (e.g. specific different core count for each node).</li>
<li><a href="ext_sensorsplugins.html">External sensor plugin infrastructure</a>
added to record power consumption, temperature, etc.</li>
<li>Improved performance for high-throughput computing.</li>
<li><a href="http://en.wikipedia.org/wiki/MapReduce">MapReduce+</a> support
(launches ~1000x faster, runs ~10x faster).</li>
<li>Added "MaxCPUsPerNode" partition configuration parameter. This can be
especially useful to schedule GPUs. For example a node can be associated
with two Slurm partitions (e.g. "cpu" and "gpu") and the partition/queue
"cpu" could be limited to only a subset of the node's CPUs, insuring that
one or more CPUs would be available to jobs in the "gpu" partition/queue.</li>
<li>Added support for native Slurm operation on Cray systems (without ALPS).</li>
<li>Added partition configuration parameters AllowAccounts, AllowQOS,
DenyAccounts and DenyQOS to provide greater control over use.</li>
<li>Added the ability to perform load based scheduling. Allocating resources to
jobs on the nodes with the largest number if idle CPUs.</li>
<li>Added support for reserving cores on a compute node for system services
(core specialization).</li>
<li>Add mechanism for job_submit plugin to generate error message for srun,
salloc or sbatch to stderr.</li>
<li>Added new structures and support for both server and cluster resources.</li>
<li>Significant performance improvements, especially with respect to job
array support.</li>
<li>Improved user support for fault-tolerance (e.g. "hot spare" resources).</li>
</ul>
</p>
<h2><a name="1403">Major Updates in Slurm Version 14.03</a></h2>
<p>NOTICE: Starting with 14.03 we will be numbering Slurm versions
with a YEAR.MONTH format.<br>
SLURM Version 14.03 release is planned in March 2014.
<h2><a name="1411">Major Updates in Slurm Version 14.11</a></h2>
<p>SLURM Version 14.11 is scheduled for release in November 2014.
Major enhancements include:
<ul>
<li>Integration with
<a href="http://en.wikipedia.org/wiki/FlexNet_Publisher">FLEXlm (Flexnet Publisher)</a>
license management.</li>
<li>Layouts framework, which will be the basis for further developments toward
optimizing scheduling with respect to additional parameters such as temperature
and power consumption.</li>
<li>Energy consumption added as a factor in fair-share scheduling.</li>
<li>Energy aware scheduling added with respect to power caps.</li>
<li>Improved user support for fault-tolerance (e.g. "hot spare" resources).</li>
<li>New partition configuration parameters: AllowAccounts, AllowQOS,
DenyAccounts and DenyQOS.</li>
<li>Scalability improvements for MPI initialization including communication of
the compute node network interface details.</li>
<li>Defer sending SIGKILL signal to processes while a core dump is in progress.</li>
<li>Other important enhancements that can not be made public at this time...</li>
<li>Communication gateway nodes to improve scalability.</li>
<li>Support for heterogeneous generic resources (i.e. user specification of
desired GPU types).</li>
<li>Support for non-consumable generic resources that are shared, but limited
in number.</li>
<li>Support for automatic job requeue policy based on exit value.</li>
<li>Add user options to set the CPU governor (OnDemand, Performance,
PowerSave or UserSpace) in addition to being able to explicitly set the
CPU frequency currently available.</li>
<li>Report Slurm message traffic by user, type, count and time consumed.</li>
</ul>
</p>
<h2><a name="1409">Major Updates in Slurm Version 14.09 and beyond</a></h2>
<h2><a name="1508">Major Updates in Slurm Version 15.08 and beyond</a></h2>
<p> Detailed plans for release dates and contents of additional SLURM releases
have not been finalized. Anyone desiring to perform SLURM development should
notify <a href="mailto:slurm-dev@schedmd.com">slurm-dev@schedmd.com</a>
to coordinate activities. Future development plans includes:
<ul>
<li>Integration with
<a href="http://en.wikipedia.org/wiki/FlexNet_Publisher">FLEXlm (Flexnet Publisher)</a>
license management.</li>
<li>Layouts framework, which will be the basis for further developments toward
optimizing scheduling with respect to additional parameters such as temperature
and power consumption.</li>
<li>Energy consumption added as a factor in fair-share scheduling.</li>
<li>Energy aware scheduling added with respect to power caps.</li>
<li>Distributed architecture to support the management of resources with Intel
MIC processors.</li>
<li>Support of I/O as a new resources, including proxy I/O nodes with data
......@@ -80,8 +69,6 @@ to coordinate activities. Future development plans includes:
post-processing, co-processing on I/O nodes, etc.) to optimize overall
system utilization.</li>
<li>IP communications over InfiniBand network for improved performance.</li>
<li>Support for heterogeneous GPU environments (i.e. user specification of
desired GPU types).</li>
<li>Fault-tolerance and jobs dynamic adaptation through communication protocol
between Slurm , MPI libraries and the application.</li>
<li>Improved support for high-throughput computing (e.g. multiple slurmctld
......@@ -93,10 +80,8 @@ to coordinate activities. Future development plans includes:
and refresh.</li>
<li>Improved support for provisioning and virtualization.</li>
<li>Provide a web-based SLURM administration tool.</li>
<li>Finer-grained BlueGene resouce management (partitions/queues and advanced
reservations containing less than a whole midplane).</li>
</ul>
<p style="text-align:center;">Last modified 05 December 2013</p>
<p style="text-align:center;">Last modified 29 May 2014</p>
<!--#include virtual="footer.txt"-->
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment
Nutzungsbedingungen | Datenschutzerklärung | Barrierefreiheitserklärung