<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"
	xmlns:content="http://purl.org/rss/1.0/modules/content/"
	xmlns:dc="http://purl.org/dc/elements/1.1/"
	xmlns:atom="http://www.w3.org/2005/Atom"
	>
<channel>
	<title>Comments on: Killing a frozen VM on a vSphere ESX host</title>
	<atom:link href="http://itknowledgeexchange.techtarget.com/virtualization-pro/killing-a-frozen-vm-on-a-vsphere-esx-host/feed/" rel="self" type="application/rss+xml" />
	<link>http://itknowledgeexchange.techtarget.com/virtualization-pro/killing-a-frozen-vm-on-a-vsphere-esx-host/</link>
	<description>A SearchVMware.com blog</description>
	<pubDate>Wed, 25 Nov 2009 06:57:30 +0000</pubDate>
	<generator>http://wordpress.org/?v=2.6.2</generator>
		<item>
		<title>By: Eric Siebert</title>
		<link>http://itknowledgeexchange.techtarget.com/virtualization-pro/killing-a-frozen-vm-on-a-vsphere-esx-host/#comment-197</link>
		<dc:creator>Eric Siebert</dc:creator>
		<pubDate>Wed, 24 Jun 2009 16:22:11 +0000</pubDate>
		<guid isPermaLink="false">http://itknowledgeexchange.techtarget.com/virtualization-pro/?p=899#comment-197</guid>
		<description>Hi Robert,

I have that procedure documented also but it seems to only work in VI3 but not in vSphere as there is no /cpu/status directory in the /proc/vmware/vm/ directory. I have it documented a bit different as shown below, if you know how to make it work in vSphere please let us know....thanks

o	Login to the service console
o	Get the vmid of the VM you want to kill by typing “vm-support –x” or “cat /proc/vmware/vm/*/names” (ie. 2533)
o	Get the world ID of the VM by typing “less –S /proc/vmware/vm//cpu/status” (substitute  with the # from the step above) Use the right arrow to scroll to the right and see the Group field value (ie. vm.2532, this is the WID of the VM), press Q to exit. A world is the software entity created in the VMkernel that runs the virtual machine. To put it another way, every Virtual Machine Monitor (VMM) has a unique world ID assigned for the duration it is powered on. It is similar to a process ID in any other operating system.
o	To kill the VM type “/usr/lib/vmware/bin/vmkload_app –k 9 ” (substitute  with the # from the step above)
o	You will see a message “Sending signal '9' to world 2532” If the command fails you will see a message “Failed to forward signal 9 to cartel 2532”, otherwise you will see not see a response.</description>
		<content:encoded><![CDATA[<p>Hi Robert,</p>
<p>I have that procedure documented also but it seems to only work in VI3 but not in vSphere as there is no /cpu/status directory in the /proc/vmware/vm/ directory. I have it documented a bit different as shown below, if you know how to make it work in vSphere please let us know&#8230;.thanks</p>
<p>o	Login to the service console<br />
o	Get the vmid of the VM you want to kill by typing “vm-support –x” or “cat /proc/vmware/vm/*/names” (ie. 2533)<br />
o	Get the world ID of the VM by typing “less –S /proc/vmware/vm//cpu/status” (substitute  with the # from the step above) Use the right arrow to scroll to the right and see the Group field value (ie. vm.2532, this is the WID of the VM), press Q to exit. A world is the software entity created in the VMkernel that runs the virtual machine. To put it another way, every Virtual Machine Monitor (VMM) has a unique world ID assigned for the duration it is powered on. It is similar to a process ID in any other operating system.<br />
o	To kill the VM type “/usr/lib/vmware/bin/vmkload_app –k 9 ” (substitute  with the # from the step above)<br />
o	You will see a message “Sending signal &#8216;9&#8242; to world 2532” If the command fails you will see a message “Failed to forward signal 9 to cartel 2532”, otherwise you will see not see a response.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Michigun</title>
		<link>http://itknowledgeexchange.techtarget.com/virtualization-pro/killing-a-frozen-vm-on-a-vsphere-esx-host/#comment-196</link>
		<dc:creator>Michigun</dc:creator>
		<pubDate>Tue, 23 Jun 2009 18:58:37 +0000</pubDate>
		<guid isPermaLink="false">http://itknowledgeexchange.techtarget.com/virtualization-pro/?p=899#comment-196</guid>
		<description>Rreynol

when I try your way, I can't find 
proc/vmware/vm/vmid value/cpu/


[CODE][root@esx41 ~]# ls /proc/vmware/vm/10804/
alloc  names[/CODE]</description>
		<content:encoded><![CDATA[<p>Rreynol</p>
<p>when I try your way, I can&#8217;t find<br />
proc/vmware/vm/vmid value/cpu/</p>
<pre>[root@esx41 ~]# ls /proc/vmware/vm/10804/
alloc  names</pre>
]]></content:encoded>
	</item>
	<item>
		<title>By: Rreynol</title>
		<link>http://itknowledgeexchange.techtarget.com/virtualization-pro/killing-a-frozen-vm-on-a-vsphere-esx-host/#comment-195</link>
		<dc:creator>Rreynol</dc:creator>
		<pubDate>Tue, 23 Jun 2009 18:02:18 +0000</pubDate>
		<guid isPermaLink="false">http://itknowledgeexchange.techtarget.com/virtualization-pro/?p=899#comment-195</guid>
		<description>Eric,

Although kill -9 is an option.  VMware support has advised, and I have experienced, that you could possibly still leave remnants of a VM running if you do that.  Here is a cleaner way to kill the VM:

1.	If the vmware-cmd does not help next up is to kill the master user world id
2.	cat /proc/vmware/vm/*/names &#124;grep vmname where vmname is the vm that is hung and find the value  for vmid
3.	less /proc/vmware/vm/vmid value/cpu/status where vmid value is the number from step above.
4.	scroll over to the right until you find the group field that shows vm.#### where the #### numbers after vm. will be the master user world id
5.	/usr/lib/vmware/bin/vmkload_app -k 9 #### where #### is the master user world id
If successful you will get a WARNING message that a signal 9 is being sent and this will cleanly kill all processes associated with the VM.

-Robert Reynolds</description>
		<content:encoded><![CDATA[<p>Eric,</p>
<p>Although kill -9 is an option.  VMware support has advised, and I have experienced, that you could possibly still leave remnants of a VM running if you do that.  Here is a cleaner way to kill the VM:</p>
<p>1.	If the vmware-cmd does not help next up is to kill the master user world id<br />
2.	cat /proc/vmware/vm/*/names |grep vmname where vmname is the vm that is hung and find the value  for vmid<br />
3.	less /proc/vmware/vm/vmid value/cpu/status where vmid value is the number from step above.<br />
4.	scroll over to the right until you find the group field that shows vm.#### where the #### numbers after vm. will be the master user world id<br />
5.	/usr/lib/vmware/bin/vmkload_app -k 9 #### where #### is the master user world id<br />
If successful you will get a WARNING message that a signal 9 is being sent and this will cleanly kill all processes associated with the VM.</p>
<p>-Robert Reynolds</p>
]]></content:encoded>
	</item>
</channel>
</rss>
<!-- dynamic -->