Backtrack:  
Blog
 
showing posts tagged with 'vmware'
 
ESXi: system logs on non-persistent storage
edited by on February 13th 2013, at 16:41
After upgrading ESXi to 5.1 on an USB device, you may get a warning message stating that the system logs are stored on non-persistent storage.



Since version 5.1, system logs are kept on the default scrap partition. If it's not specified, it defaults to ramdisk, which is non-persistent; upon rebooting, the system logs will be cleared.

To resolve the issue, you can manually specify a location for the logs to be stored. This can be any configured datastore (VMFS or NFS). To configure this, go to the configuration of your ESXi host, Software, Advanced Settings.



In the tree on your left, find the Syslog, global configuration section and enter a datastore name in the field Syslog.global.log  ...
 
vmwaretray DMP files in user profiles
edited by on November 14th 2011, at 15:29
If you're running RDS on a vSphere server with VMware Tools installed, you may have noticed the evergrowing presence of vmwaretray DMP files in the user profiles. These are crash dumps from the VMware Tray icon, which is ran during every session.

The reason for this crash dump is an error which is caused by lack of read access to a certain key in the Windows registry by your regular (domain) users.

This key is called HKEY_LOCAL_MACHINE\Software\VMWare. Inc.\VMWare Tools.

Set the permissions on that key to read access for the users logging on to the server, and your problem should go away.

Hint: you may also wish to remove the icon altogether. This is done by setting a certain registry va  ...
 
Remove VMware Tools icon from system tray
edited by on November 14th 2011, at 15:28

If you're running RDS on a vSphere server with VMware Tools installed, you may wish to remove/hide the system tray icon. This can be done by setting a certain registry value.

The registry value is:

  • Key: HKEY_LOCAL_MACHINE\SOFTWARE\VMware, Inc.\VMware Tools
  • Value: ShowTray

Change the value to 0 will effectively disable the icon the next time a user logs on.

 
Cannot delete data store of freshly installed ESXi
edited by on April 4th 2011, at 15:51
With the installation of VMware ESXi 4, data stores will automatically be created using the default settings. This includes a block size of 1 MB. If you don't want that, you could delete the empty datastore and create a new one.

Sometimes, deleting the empty datastore fails with the message:

Quote
Call "HostDatastoreSystem.RemoveDatastore" for object "ha-datastoresystem" on ESXi "xx.xx.xx.xx" failed.

VMware's KB 1017104 provides the solution for another problem, but this works just as well with this issue, with a little variation.

Notice
Performing the steps in this solution will totally utterly destroy your data store. Do not use it if there's still any req  ...
 
Install VMware video driver in Windows 2008R2
edited by on March 7th 2011, at 15:22

In VMware ESXi 4.1, by default, the video drivers are not correctly installed. This is because of the fact that the old driver (which is installed by default) is a WDM type of driver. This type of driver has been kicked out of Windows 7 and Windows 2008R2 in favour of the newer WDDM driver architecture. As a result, the video is reverted to the Standard VGA adapter, which has terrible performance.

However, installing the VMware Tools does provide a correct driver, even though it is not installed by default. To install it, see the second chapter of the article: Windows 2008 R2 console freezes in vSphere console.

 
Windows 2008 R2 console freezes in vSphere console
edited by on October 15th 2010, at 11:54
At work, I'm currently deploying a Windows 2008 R2 on a VMware ESXi 4.0, and noticed an odd behaviour. At irregular intervals, the console of the guest OS simply locks up and doesn't do anything anymore. Networking and remote access seem to work fine, it's just the console that is freezed. Restarting the vSphere client doesn't help, and the only way to get rid of the problem is by hard-resetting the VM. The issue only occurs on Windows 2008 R2, the other VMs run fine.

The problem is caused by the VMware SVGA II driver, and has two underlying issues: the driver itself, and an apparent lack of video RAM (due to the new HW-accelerated GDI of 2008R2/Win7).

The most quickest way is to get rid o  ...
 
Convert thick disks to thin provisioning
edited by on September 13th 2010, at 15:05

It is not possible to convert disk files from thick to thin, in-place. Instead, cloning the disk (or machine) allows you to change the provisioning type and thus enabling to have a thin disk. After the clone is complete, delete the original and you're good to go.

You can either clone a complete machine, or just clone a disk using the CLI or service console:

vmkfstools -i <srcDisk.vmdk> -d thin <dstDisk.vmdk>
 
Consolidate snapshots from PowerCLI
by on January 1st 1970, at 01:00

Occassionally, it may be necessary to perform disk consolidation. If you have a lot of VMs which need consolidation, it can be tedious to do this in the webclient. Fortunately, it's also possible to mass-consolidate via PowerCLI.

Install and log in using PowerCLI.

To show which VM's need consolidation, run:

Get-VM | ? {$_.Extensiondata.Runtime.ConsolidationNeeded}

To actually perform disk consolidation, run:

Get-VM | ? {$_.Extensiondata.Runtime.ConsolidationNeeded} | % {$_.ExtensionData.ConsolidateVMDisks_Task()}
 
Prevent VMware Tools auto-update in Fusion
by on January 1st 1970, at 01:00
By default, VMware Tools are automatically updated when power-cycling a VM each time a new version is available. This may not be desirable, and currently, there is no way to change this behaviour anywhere in the settings. This setting can however by changed manually on a VM-basis, by manually editing the VM's configuration file (.vmx).

Although there's mention of a global setting in Fusion's configuration files, many users report this as not working, so I only mention the per-VM setting here.

First, make sure your VM is not running. Preferably, close VMware Fusion entirely (⌘ + Q).

Using Finder or Terminal, navigate to the VM folder. When using Finder, you'll need to right-click you  ...
 
Enable/disable automatic storage reclamation (VMFS unmap) via PowerCLI
by on January 1st 1970, at 01:00
In VMware, using VMFS-6 as the datastore filesystem introduces automatic storage reclamation. This process, also called "VMFS unmap" runs in the background (by default at low priority so it doesn't interfere with performance) to clear allocated unused storage space by sending "unmap" commands to the underlying storage so it can reclaim the blocks for other uses.

Sometimes it may become necessary to (temporarily) turn off (or on, if it's turned off already) storage reclamation on a datastore. You can do this through the WebClient but if you want to change this for a bunch of datastores, you can also do this via PowerCLI. While the latter is pretty complicated, William Lam  ...
 
Error java.lang.reflect.InvocationTargetException when joining vCenter to AD
by on January 1st 1970, at 01:00
When attempting to join your vCenter appliance to AD, you may get the following error message when clicking the Join button:

Error
java.lang.reflect.InvocationTargetException

This is known behaviour when running vCenter with an external Platform Service Controller (in HA mode) but the issue may persist even after upgrading vCenter to a new version with the option to switch to an embedded PSC. You can work around the issue by performing the domain join from the CLI (through SSH):

Connect to the vCenter appliance through SSH (if it's still an external PSC, connect to that instead).

Unless shell access is already activated, use the shell command to enter the regular shell.

Run the followin  ...
 
VMs in Already Used status in VMware Horizon View
by on January 1st 1970, at 01:00
In VMware Horizon View VDI, when using floating pools with instant clones, you may run across a problem where VMs get the "Already Used" status after users log off. The VM remains unavailable for further use and does not get deleted or refreshed.

The reason for this usually is that some process is blocking the refresh/delete operation, causing the VM to remain unusable and powered on to allow a system administrator to investigate the issue. If this is not desired (e.g., the reason for the blockage is known but cannot be resolved immediately), there is a policy which can be set to allow for automatic clean up of these "dirty" VMs.

The attribute is called pae-DirtyVmPolic  ...
 
 
showing posts tagged with 'vmware'
 
 
« ‹December 2024› »
SunMonTueWedThuFriSat
1234567
891011121314
15161718192021
22232425262728
293031    
 
Links
 
Quote
« When a bird does poo poo in your eye, be happy elephants don't fly. »