diff --git a/NEWS b/NEWS index 176b87a6d793c0151cfeb4a7855dabc6e7b60eda..468f9c967551216ea7c12055761e294c5ed90f9e 100644 --- a/NEWS +++ b/NEWS @@ -22,6 +22,7 @@ documents those changes that are of interest to users and admins. WARNING: "JobAcctType=jobacct/log" is no longer supported. -- Removed vestigal 'bg' names from bluegene plugin and smap -- InactiveLimit parameter is not enforced for RootOnly partitions. + -- Update select/cons_res web page (Susanne Balle, HP, cons_res_doc_patch_3_29_06). * Changes in SLURM 1.1.0-pre2 ============================= diff --git a/doc/html/cons_res.shtml b/doc/html/cons_res.shtml index 2dc4e3437b8e0e603d468056e94dfb4c42ae4439..b6f9691aed06b5e56d8460cac3887cd9f6e53968 100644 --- a/doc/html/cons_res.shtml +++ b/doc/html/cons_res.shtml @@ -35,6 +35,16 @@ this plug-in is described below. view, working the same way as when using the default node selection scheme.</li> <li>We introduce a new switch <i>--exclusive</i> which will allow users to reserve/use nodes in exclusive mode even when consumable resources is enabled. see "man srun" for details. </li> + <li>Using the <i>--overcommit</i> or <i>-O</i> switch in the + <b>select/cons_res</b> environment is only allowed/possible when a user + requests dedicated nodes using the <i>--exclusive</i> + switch. Overcommiting CPUs in a non-dedicated environment would + impact jobs that are co-located on the nodes which is not a + desirable feature.We are currently looking into "pinning" the + processes to jobs' allocated CPUs which should allow users in + <b>the future</b> to overcommit CPUs in a the <b>select/cons_res</b> environment + without affecting co-located jobs.</li> + <li>SLURM's default <b>select/linear</b> plugin is using a best fit algorithm based on number of consecutive nodes. We have chosen the same node allocation approach for consistency.</li>