Abbott Internet Consultants

Number one cause of HA errors in VMware is incorrectly configured FQDNs.  If you find that HA is throwing up an error because you forgot a 03 and used 3 for example (hypothetically),  then you need to correct this to get it all working.

The hostname command should do the trick, but you’ve got a couple of other places to look in as well.  This should all be so familiar to anyone used to Red Hat anyway.

Make sure it all matches up in:

/etc/sysconfig/network

/etc/hosts

You may also need to reboot, sometimes a service network restart is not enough.

HA should now work a treat!

Increase the size of the disk presented to the VM using vCenter, then go to disk manager on the VM itself.  You should now see some unused space.

Launch diskpart from the command line.

list volumes

select volume #

extend

Done!

Get all VMs using a certain datastore (in this case one with test in the name), and then show those that are powered off.

[vSphere PowerCLI] C:Program FilesVMwareInfrastructurevSphere PowerCLI> get-datastore -name "*test*" | get-vm | wher

e { $_.Powerstate -eq "poweredoff" }

Name                 PowerState Num CPUs Memory (MB)

----                 ---------- -------- -----------

UPRINTXDM            PoweredOff 1        2048

Uprint8_test         PoweredOff 1        512

U-SANTEST-01         PoweredOff 1        512

U-SANTEST-02         PoweredOff 1        512

TestVM               PoweredOff 1        512

uclh-test-vm02       PoweredOff 1        3840

Get all hosts in a cluster to rescan for new storage:

Get-Cluster | get-vmhost | Get-VMHostStorage -RescanAllHBA

How to reduce the migration time on a datacenter move to the absolute minimum?  Powershell helps hugely with this as there are a large number of mouse clicks required to re-import the virtual environment.  More details on the move to follow, but for now lets get started.

Go and install powershell on your windows machine, and then go to the vmware site to download their VI-plugin.

Once installed, you can launch from the vmware folder group, but you may get an error when launching due to the unsigned script.  You can ignore this but it may cause you further issues later.

If you get this you’ll see this:

File C:\Program Files\VMware\Infrastructure\vSphere PowerCLI\Scripts\Initialize-VIToolkitEnvironment.ps1 cannot be load
ed because the execution of scripts is disabled on this system. Please see "get-help about_signing" for more details.
At line:1 char:2
+ . <<<<  "C:\Program Files\VMware\Infrastructure\vSphere PowerCLI\Scripts\Initialize-VIToolkitEnvironment.ps1"
PS C:\Program Files\VMware\Infrastructure\vSphere PowerCLI>

To resolve, you need to set the signing to allow the VMware bits to all run properly.  Most cmdlets will be fine, and I only ran into problems with this when I tried to use the AnswerVM class which didn’t exist.

Note: this will remove your security on running scripts, so only do this if you are comfortable with the implications - don’t let any code run you’re not happy with and understand fully.

set-executionpolicy unrestricted

Then exit the shell, and restart - it will now look like this:

VMware powershelll startup

Sometimes it is the blindingly obvious that gets you.  After running into issues with creating a mount point for ISO images to be used by the host servers, we ran into a couple of problems.

Firstly, -t smbfs now appears to be deprecated, so -t cifs should be used instead.

Secondly, smbclient traffic is blocked by the firewall by default, and all you will get when trying to make the connection is:
mount error 111 = Connection refused

To resolve this, simply go into the configuration for the host, security profile and make sure to enable smbclient.

First installation of production ESX4 is almost completed now for a client in Washington DC.  I’ll be adding more notes documenting the following:

• general installation and configuration
• iSCSI & EqualLogic setup and performance tuning
• Alerting and management
• Troubleshooting

I ran a virtualisation seminar yesterday for our friends at NCC and some of their clients.  Covering topics from “what is virtualisation?” to a full discussion of the in and outs of handling the organisation change that it brings, I think it gave a very good grounding for the attendees in how to build a business case for virtualising their data centres, avoiding some of the pitfalls.

We’re running a further workshop in London this time, more details and registration can be found at the NCC Website.

For all the move to cloud/SaaS it seems like Google have really dropped the ball a little, with a breach to their docs service.  Now I use Google docs as a convenient way of sharing bits of information, but you’ve got to wonder about any corporate willing moving from internal MS Word/Excel to Google until the security is not an issue.

http://www.theregister.co.uk/2009/03/09/google_docs_serious_security_breach/

I presented this week at the NCC round table event on “virtualisation”.  There was a good mix of attendees, from finance, manufacturing, local government and the service sector.  All organisation were using VMware already to one extent or another, but still in a fairly limited manner.

The biggest surprise was probably that so few were using DRS and HA in their production environment.

Lots of interest about the upcoming features in ESX 4.0 however, especially the lockstep VM/data protection feature.