diff --git a/doc/html/slurm_ug_agenda.shtml b/doc/html/slurm_ug_agenda.shtml index 6239bb733f0658cbd00e6771f12e02766226965a..5e4c05c190517172bf72dadcf75a4bc2ac948ced 100644 --- a/doc/html/slurm_ug_agenda.shtml +++ b/doc/html/slurm_ug_agenda.shtml @@ -254,13 +254,19 @@ plus dinner on Wednesday evening.<br><br> </tr> <tr> - <td width="15%">15:10 - 15:40</td> + <td width="15%">15:10 - 15:30</td> + <td width="15%"> Technical</td> + <td width="25%"> Francois Daikhate, Matthieu Hautreux (CEA)</td> + <td width="45%"> Depth Oblivious Hierarchical Fairshare Priority Factor</td> +</tr> +<tr> + <td width="15%">15:30 - 16:00</td> <td width="15%"> Technical</td> <td width="25%"> Morris Jette (SchedMD), Yiannis Georgiou (Bull)</td> <td width="45%"> Slurm Roadmap</td> </tr> <tr> - <td width="15%">15:40 - 16:30</td> + <td width="15%">16:00 - 17:00</td> <td width="15%"> Discussion</td> <td width="25%"> Everyone</td> <td width="45%"> Open Discussion</td> @@ -577,6 +583,17 @@ GPUs to specific CPUs.</p> <p>This site report will detail the recent introduction of Slurm on a new computer at Technische Universitat Dresden.</p> +<h3>Depth Oblivious Hierarchical Fairshare Priority Factor</h3> +<p>Francois Daikhate, Matthieu Hautreux (CEA)</p> +<p>As High Performance Computing use becomes prevalent in increasingly varied +scientific and industrial fields, clusters often need to be shared by a growing +number of user communities. One aspect of managing these heterogenous groups +involves being able to schedule their jobs fairly according to their respective +machine shares. In this talk we look at how slurm hierarchical fairshare +algorithms handle this task when user groups form complex hierarchies. We +propose an alternative formula to compute job priorities which improves +fairness in this situation.</p> + <h3>Slurm Roadmap</h3> <p>Morris Jette (SchedMD), Yiannis Georgiou (Bull)</p> <p>Slurm continues to evolve rapidly, with two major releases per @@ -587,6 +604,6 @@ computing. These requirements include not only scalability issues, but a new focus on power management, fault tolerance, topology optimized scheduling, and heterogeneous computing.</p> -<p style="text-align:center;">Last modified 3 September 2013</p> +<p style="text-align:center;">Last modified 5 September 2013</p> <!--#include virtual="footer.txt"-->