Differences

This shows you the differences between two versions of the page.

--- cluster:87 [2010/05/13 13:36] – hmeij
+++ cluster:87 [2010/05/19 20:55] (current) – hmeij
@@ Line 2: / Line 2: @@
 **[[cluster:0|Back]]**
-==== RFQs ====
-The Steering Committee has decided to break the purchase of the next cluster up into three pieces.  Partly because we're not at capacity in our current cluster environment.  The components are:
+==== Cluster Software ====
-  * Head Node, Disk Array, UPS, and compute nodes to fill Rack #1 ($125,000)
+  * A good overview, swallowtail is platform OCS/LSF and sharptail is Kusu/Lava
-  * A Database Server (can go to either rack) ($25,000)
-  * Compute nodes filling Rack #2($100,000)
-The dollar amounts are guidelines and not fixed.  The intent is for this new cluster to become our primary cluster and potentially ingest our old clusters for a single point of access/job submissions.  At this time, we are requesting quotes for Rack #1.  However, if you wish to submit all three quotes now, that would be fine too.
+  * [[http://www.hpcprojects.com/features/feature.php?feature_id=235|HPCprojects]]
-==== Details ====
+  * And I guess this is the spinoff of Kusu/Lava - UniCluster (scheduler?)
-===== Head Node =====
+  * [[http://www.unicluster.org/|UniCluster]] ... works with schedulers ???
+  * Then there is HP's version [[http://h20311.www2.hp.com/HPC/cache/412128-0-0-0-121.html?jumpid=ex_r1459_w1/en/large/tsg/go_cmu#overview|CMU]]
+  * And the Sun/Oracle [[http://gridengine.sunsource.net/|Grid Engine]]
+  * And MOAB ... works with PBS, Torque, LSF, Grid Engine
+  *
 \\
 **[[cluster:0|Back]]**