Differences

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

--- cluster:132 [2014/08/08 16:06] – hmeij
+++ cluster:132 [2014/08/11 14:08] (current) – [Tests] hmeij
@@ Line 30: / Line 30: @@
-Ah, I'm simply going to CentOS7 because LXC is fully integrated
+Ah, I'm simply going to CentOS-7 because LXC is fully integrated
   * [[http://www.theregister.co.uk/2014/07/07/centos_7_gm/]]
@@ Line 37: / Line 37: @@
 ==== KVM ====
-KVM is hardware level virtualization as opposed to my previous meanderings into the world of software level virtualization (Xen).  There is basically a "dom0" domain manager running keeping track of virtual images that run from block devices written onto a block device.
+KVM is hardware level virtualization as opposed to my previous meanderings into the world of software level virtualization (Xen [[cluster:127|Virtual HPCC services]]).  There is basically a "dom0" domain manager running keeping track of virtual images that run from block devices written onto a block device.
   * Followed the primary setup on this page
@@ Line 48: / Line 48: @@
     * http://rwmj.wordpress.com/tag/virt-clone/
-First I build my v1 clone the way I wanted it from ISO and then customized it manually (/etc/fstab, openlava scheduler, /etc/passwd, /etc/shadow, /etc/group and /etc/hosts, etc). Added a second bridge (''br1'' is public/internet, previously created ''br0'' is for private network for scheduler) later via virt-manager GUI (view the details of vm and then select Add Hardware). ''br1'' allows me to run yum update before I clone. Power this clone up, add to a test queue, and submit some jobs to make sure it all works.  Then clone.
+First I create my v1 clone with virt-install then build my v1 clone the way I wanted it and then customized it manually (/etc/fstab, openlava scheduler, /etc/passwd, /etc/shadow, /etc/group and /etc/hosts, etc). Added a second bridge (''br1'' is public/internet, previously created ''br0'' is for private network for scheduler).  That was done via virt-manager GUI (View VM Details of vm and then select Add Hardware). ''br1'' allows me to run yum update before I clone. Power this clone up, add to a test queue, and submit some jobs to make sure it all works.  Then clone.
 <code>
@@ Line 62: / Line 62: @@
 </code>
-Here are the steps for cloning v1 to v5. You'll need ''guestfish' part of the libguestfs tools allowing you to edit content inside the block level file.  Very nifty.
+Here are the steps for cloning v1 to v5. You'll need ''guestfish'' part of the libguestfs tools allowing you to edit content inside the block level file.  Very nifty.
   * yum install libguestfs-tools-c
-Next we'll create the v5.img block device, dump the v1 config into v5.xml and then edit that file. UUID and Mac Address we'll edit and the last 2 characters we change to '05' (make it unique). Name becomes v5 from v1 and we'll similarly adjust the block device location/filename. Then we define this as domain v5 vm.
+Next we'll create the v5.img block device, dump the v1 config into v5.xml and then edit that file. UUID and Mac Address we'll edit and the last 2 characters we change to '05' (make it unique). Name becomes v5 from v1 and we'll similarly adjust the block device location/filename. Then we define this as domain v5.
-Launch ''guestfish' and make the same edits to Mac Address in files listed below. Then boot the vm.
+Launch ''guestfish'' and make the same edits to Mac Address in files listed below. Then boot the vm.
 <code>
@@ Line 95: / Line 95: @@
 ><fs> edit /etc/sysconfig/network-scripts/ifcfg-eth0
 ><fs> edit /etc/sysconfig/network-scripts/ifcfg-eth1
-(for nics comment out any UUID lines, change hardmare address, change IP)
+(for nics comment out any UUID lines, change hardware address, change IP)
 ><fs> edit /etc/udev/rules.d/70-persistent-net.rules
 (change hardware address)
@@ Line 107: / Line 107: @@
 ==== Tests ====
-<code>
-   4   6   8   10  12  14   16
-560 588 625 760 936 1122 1300
- (linear)       781 938 1094 1250
- (n35:load 32)  726
-<code>
+^ Melt Lammps LJ problem, 10000 steps with 32000 atoms Loop Times (secs) ^^^^^^^^^
+|  2  |  4  |  6  |  8  |  10  |  12  |  14  |  16  |  nr of jobs  |
+|  556  |  560  |  588  |  625  |  760  |  936  |  1122   |  1300  |    |
+|  (linear)  ||||  781  |  938  |  1094  |  1250  |    |
+|  (n35:load 32)  ||||  726  |    |    |    |    |
+|  (hyperthreading)  ||||    |    |    |  ????  |  16 logical cores  |
+VMs running in queue ''testvm'' on dual quad core PE2950 (8 cores, 16 GB ram)
+  * First ran on n35 (32 core node under hyperthreading will full load) with an average loop time 726 secs.
+  * As I submit jobs to the VMs they perform well up to 8 jobs (one job per core; dual quad core node).
+    * That is with the KVM overhead
+  * Assuming a linear penalty for over committing, 16 jobs is expected to take Loop times of 1250 secs.
+    * However after 10 jobs we're surpassing that penalty threshold
+  * And then I was to turn on Hyperthreading creating 16 logical cores
+      * To my dismay this chipset does not support that, bummer!
+      * Was expecting to gain some performance back ...
+      * Maybe try on newer hardware when idle ...
+But we learned Xen and KVM setups, and
+  - We can now support a heterogeneous environment if we wanted (Suse, Scientific Linux, Windows (eh, what?))
+  - Use a KVM environment up to the number of cores on a box without penalty
+  - And change the mix of nodes if needed (more/less cores per node, memory size etc)
+  - Still, not an answer for my "high core count/low memory footprint" problem, see [[cluster:133|high core count/low memory footprint]]
 \\