Think IPM

Wednesday, April 30, 2008

Enabling DRS

DRS or Distributed Resource Management allows administrators to schedule automated vMotion based on load requirements in the farm.

Under the properties of the Cluster, enable DRS.


Once Server is in a cluster, you can set the migration threshold.

Setting DRS Rules

DRS rules allow administrators to keep or separate virtual machines. The two rules allowed are keep together for machines that are dependent on each other and keep apart for machines that are redundant to each other.

Click Here to Continue Reading >>

VMware Tools Installation

Just some random documentation I had laying around on the VMware Tools Installation on a Windows Machine.

VMware tools allow for a smoother user experience. Tools consist of an updated Mouse, Graphics, memory and NIC drivers specifically optimized for VMware.

From the VI Client, run VMware Tools installation

Reboot Server

Click Here to Continue Reading >>

Enabling NTP on ESX Servers

Time Notes

Time drift within Virtual machines is very common. Time is calculated within Operating Systems based on the assumption that it has 100% access to the CPU0 for timing. In a shared Virtual Environment, that is not the case and can cause time drifts in relation to CPU workload.

If your Time Sources are Virtual Machines, the Time Sources (i.e. Domain Controllers) should not be set to sync time from the hosts via VMware Tools.

The NTP client on each ESX host should be set to a physical Time Source. If none are available within the organization, round robin DNS records have been configured to gather time from the internet via UDP port 123.



Click Here to Continue Reading >>

Microsoft Operating System Licensing under VMware

For those you with Microsoft Windows 2003 R2 licenses and some sort of virtualization (Xenserver or VMware), you might want to double-check to make sure you are not over allocating your licenses. Under Microsoft's R2 licensing, 1 purchased copy of Windows 2003 R2 Enterprise entitles you to run 4 virtual machines. If you choose to virtualize the 38 Windows 2003 Enterprise servers and upgrade to the R2 licensing, you could potentially be licensed for an additional 115 virtual servers.


More information on Microsoft's R2 licensing can be found here: http://www.microsoft.com/licensing/highlights/virtualization/faq.mspx


Click Here to Continue Reading >>

Tuesday, April 29, 2008

Patching an ESX server with Update Manager

Update Manager

VMware Update Manager
automates patch and update management for VMware ESX Server hosts and virtual machines. Update Manager addresses one of the most significant pain points for every IT department: tracking patch levels and manually applying the latest security/bug fixes. Patching of offline virtual machines enforces higher levels of patch standards compliance than physical environments. Integration with VMware Distributed Resource Scheduler (DRS) enables zero-downtime VMware ESX Server host patching capabilities.

Host Patching

This process details the Patching of an ESX host. The remediation process involves Migrating Virtual Machines off the host using vMotion, patching and then redistributing workloads via DRS.

After connecting to Virtualcenter,

you must enable the Update Manager plugin

Download and install the Plug-in

Verify that the plugin is ENABLED.

Once enabled, a new task button and Tab will be available.

The next step is to attach a baseline to the Cluster object (Which holds the Hosts)



Baselines are either Fixed (which you manually select which updates get applied) or dynamic (where updates are applied as they become available from the vendors)


The baselines are further divided into Non Critical and Critical classifications.

Once the baselines are attached you can SCAN FOR UPDATES.


Once the Scan has been completed and Update Manager has downloaded the needed updates, You can choose to REMIDIATE a server.


I would recommended that you only do 1 host at a time.

Choose the baselines or updates you would like to apply and click Next.



After clicking Finish, the Host will go into Maintenance mode, vMotion the VMs to other hosts, Patch itself, reboot and then return to normal operation.

Click Here to Continue Reading >>

VMware Distributed Power Management

Enabling DPM on VMware 3.5 / Virtualcenter 2.5

DPM or Distributed Power Management is an experimental feature that in conjunction with DRS [Distributed Resource Scheduling], will migrate machines off of underutilized hosts in an effort to put the hosts into standby mode to conserve energy. As resource demand increases, 'Magic Packets' will be sent to the hosts to wake them from standby mode and Virtual Machines will be migrated back onto the available hosts.


Currently, I recommend only setting this to Manual since DPM is not fully support by VMware. Another issue I have with it is the inability to set the lower threshold for ON servers. Without this, DPM will shutdown all servers except one if resources permit. This obviously is not good for production since you will no longer have fault tolerance with only one server.

From a conceptual standpoint though, this is a great feature to have. I hope VMware continues to develop it in future releases.


Click Here to Continue Reading >>

Remotely Granting Shell access to Root

Grant Shell access to Root

By default, ROOT no longer has the ability to SSH into the ESX server.
This can be re-enabled for ease of administration. If you are at the console, skip directly to the PUTTY commands.

Add Service Account User via VI Client

Add Temp Service Account with Shell Privilege.

PUTTY to 172.27.32.111


Log in as our temp VMService account and type su - and the ROOT password.

Type nano –w /etc/ssh/sshd_config

Find PermitRootLogin no and add #

Restart sshd by typing service sshd restart

Remove Temp VMService Account

Click Here to Continue Reading >>

V2P Part II

Fixing a BSOD after an V2P using BartPE

If you have just V2P'd a server from Virtual to Physical and the server is now experiencing a Blue Screen of Death upon Reboot, You most likely do not have the correct Mass Storage Device installed. VMware Virtual Machines are configured to use VMware SCSI Adapters for their storage. When you V2P them to an HP server, Windows requires that it has a Smart Array controller to correctly boot.

You could pre-install the driver into the Virtual Machine and reimage OR you could use BartPE and it's great plugins to fix the physical box now.

Obtain your copy of BartPE @ http://www.nu2.nu/pebuilder/

Make sure that your copy of BartPE has the FixIscsi Plug installed.
http://www.rtfm-ed.co.uk/downloads/fixvmscsi-xpe.zip

Additional information can be found for this plugin @
http://www.rtfm-ed.co.uk/downloads/whitepaper-ultimatep2v-updates.zip

 

After V2P'ing the image back down to the physical server, power on to ensure that the image restoration was completed successfully.

 

If you are able to see a BSOD, you have successfully restored the image. The BSOD occurs because the Mass Storage Device Driver is set incorrectly and Window's cannot find its internal hard drives.

 

BartPE and SCSIFix plugin to the rescue.

 


Boot CD into BARTPE SCSIFix CD.

 

Select NO for networking support. This is not needed to correct the BSOD.

Select the Fix-VMSCSI program and Set Disk0 to be C.

 

Set OS Target Root to C:\

 

After a short command script window executes, return to select the appropriate Storage Driver for the physical box.

You can also use this feature to select IDE which is appropriate for Citrix XenServer Images.


After selecting the driver, a small command script window will execute.

 

You may be prompted for the TARGET_ROOT path. Enter in C:\Windows [ENTER]

 

After the Mass Storage Device Driver insertion is completed, restart the Physical Server and Windows should boot normally.
Plug and Play will kick in and recognize and install other newly found hardware.
Allow this Windows process to complete uninterrupted and reboot as necessary.

Click Here to Continue Reading >>

Monday, April 28, 2008

Using Acronis to take an image of a Virtual Machine [V2P Part 1]

Here are some straight forward instructions on how to take an Acronis image of a Virtual Machine. This is part 1 of a P2V [Physical to Virtual] migration of a machine.

Taking an Image of the Virtual Machine

Log into VirtualCenter and boot Server to Acronis CD.


To boot a VM to Acronis, you must have an Acronis CD or ISO available to you.


After obtaining the media, reboot the Virtual Machine to the CD or ISO.


Choose Acronis True Image Enterprise.

Choose Backup. This option will be used to create an 'Image' of the Current VM. You will need to have enough space on the network to store the resulting TIB file.



Choose Entire Disk Contents for inclusion of the image.


Choose the Disk. If this were a physical server, you might see additional partitions such as utility partitions that you would deselect.


This screen also gives you an idea of how much space you will need on your destination drive.

Connect to your destination Share.


Be sure to WAIT while typing in the path to allow the system to prompt you for credentials.


Once Authenticated, press next to begin Imaging process.

Choose Create New Full Backup Archive.


Choose Next to continue Imaging process.

No additional settings are required for Virtual Machine migrations.

Verify all information and Click next to begin imaging process.


Once the imaging process is completed, you can restart the server or leave it shut down and begin the restoration process.

Click Here to Continue Reading >>