We are 100% Virtualized, what next?

This is the question every organization has who have achieved their goal to virtualize their physical workloads either on vSphere or any other Hypervisor. From here on the journey becomes interesting and from this point the organizations start thinking about Cloud. So what exactly is Cloud Computing? Well if you have to define it in one word then it’s basically “Standardization”. Cloud Computing should be able to deliver on demand workload provisioning and should be accessible over the internet, well there can be many definitions for Cloud Computing, but I would like to keep things simple and clear to understand. I will not go more in details about Public and Private Cloud as those terms are pretty much self-explanatory and by now everyone in the virtualization industry knows about it.

Let us discuss more about what do we need to implement a successful cloud platform? In today’s world the user requests their IT admins to provision the Server, although the provisioning is fast and requires less effort compared to provisioning a physical server, but then,  it still requires the IT admin to provision the Virtual Machine and the backened Compute and Network resources manually. So how do we automate these tasks? Simplify provisioning? And enable the end user to provision their own workload without depending on their IT administrators?

One simple answer is “Orchestration and Automation”, but again, we are in the midst of transformation where there are so many tools available in the market and it’s hard to decide which one best suits your requirements. It’s just not technical aspects of your infrastructure but the ITSM model which adds to decision making.

I have had the opportunity to work with different automation tools which basically provide a self-service portal where the end user has a choice to order a service, may it be a Virtual Machine or any value added service, in an essence the goal is to provide “Anything as a Service”.

Let me share some highlights of different automation tools, I am not going to deep dive but just a quick comparison to get some understanding:

Cisco Intelligent Automation Center by Cisco

Cisco IAC allows as end user to order what they need from a Service Catalog with standard definitions. The deployment architecture is very simple where we have a cloud portal application server and an Orchestration Server which makes all the web service and API calls to all the Virtualization element managers (vCenter, UCS manager etc.). Cisco IAC can provision Virtual Machines as well as bare metals. Cisco Intelligent Automation for Cloud orchestrates resource-level operations across compute resources such as Cisco Unified Computing System (Cisco UCS) or other hardware; hypervisors such as VMware, Xen, or Hyper-V; storage resources such as EMC and NetApp; and network resources such as the Cisco Nexus® family. Cisco IAC Multicloud Acceleration kits allow you to integrate with third party cloud portal such as vCloud Director, Amazon EC2, OpenStack etc.

In today’s context, most cloud platforms either leave out Network services or haven’t come up with an adapter which tightly integrates with your core networking platform. With Cisco IAC, the Network Services Manager enable customers to order their networking resources at the same time they order their virtual data center. An NSM adapter permits rapid building of NSM automation for network provisioning. Network automation requirements can be customer specific, completion of the automation process in the customer’s back end processes is completed by trained staff or through a services engagement.


Cloupia is now part of Cisco, Cloupia Unified Infrastructure Controller or better known as CUIC is a self-service portal which can interact with Converged Infrastructure like Cisco UCS, NetApp’s FlexPod and provide IaaS to the end users. CUIC can provision virtual machines as well bare metal servers. It can manage compute, network, Storage and Virtualization stack from single pane of glass. CUIC has Physical Connectors that talks to Converged Infrastructure, Virtual Connectors that talks to VMware vCenter, Hyper-V, Red Hat etc and Cloud Connectors that talks to Amazon EC2, Rackspace etc. It even provides a mobile application “CloudGenie” which can be loaded on your IOS or Android devices.

The deployment architecture is very easy, you just need to deploy an OVF appliance with a supported database and you are good to provide IaaS to your end users with a well-defined Self Service Catalog

vCloud Director by VMware

vCloud Director provides IaaS to the end users via a self-service web portal. The cool VXLAN features allows an end user to provision as many networks as they want. It can also provide services like firewall, VPN, Static routing etc. The only drawback of vCloud Director is that it only integrates with vSphere and no third party hypervisors, also, it doesn’t provide the ability to provision bare metal servers.

The deployment model is straight forward, just need to setup a linux VM and install the vCloud Director binaries and then onboard the resources from your vCenter Servers

I haven’t really had a chance to deep dive with vCloud Automation Center, but the product looks promising and very soon I am planning to get my hands dirty on it.

BMC Cloud Life Management:

The BMC CLM is another Self Service portal which allows users to order services from standard catalog items. It supports integration with VMware, Amazon, OpenStack and Hyper-V.

vCenter Chargeback Manager Quick bites

The vCenter Chargeback Manager is used to provided Chargeback and Showback reporting of vCenter and vCloud Director workloads. The current version is 2.5.0 and can be installed on W2K3 R2 and W2K8 32 R2 (32bit and 64 bit versions) and requires SQL or Oracle as their backened database. Since the front end is web based, it requires IE/Mozilla and Adobe Flash Player.

Data Collectors:

– Chargeback data collector connects to vCenter Server to gather metrics for Virtual Machine usage

– vCloud Data collector connects to vCloud director database and monitors all the vCD chargeback-related events. It creates hierarchy based on Organization

– vShield Manager data collector gathers information about vCloud Networks, Services like NAT, Firewall, VPN etc…

– Additional Data collectors can be installed for increased availability, they act in an active/passive manner

Kendrick Coleman has written an excellent step by step blogpost here to install vCenter Chargeback


– Reports can be run/scheduled and emailed from the Chargeback UI and can be exported in PDF, XLS and RTF formats

– Chargeback APIs allow ability to export reports in XML format so that it can integrated into custom billing solutions

– Chargeback manager can generate Cost, usage and comparison reports for hierarchies and entities

Multisite deployment and Clustering:

– vCenters from multiple sites can be added to vCenter Chargeback Manager

– Maximum of 10 vCenters are allowed

– Only one instance of vCenter Chargeback can be registered as a plug-in for a single vCenter Server

– Chargeback Manager can be deployed as a cluster to serve load balancing and High Availability. Although all the Chargeback instances communicate with a single shared database

– Chargeback uses internal load balancer to distribute the load across chargeback instances via Round Robin

– Multicasting must be enabled on the machines running each Chargeback instance

Security and Permissions:

– Charge manager has pre-defined roles with Super User having absolute power,

– Administrator has access to only users/resources and resources created by users that he created

Configuration required within Chargeback Manager:

– Every organization in the vCloud Director is created as a hierarchy via API. The hierarchy includes all Org vDCs, media and template files, vApps, VMs and networks

– Can create custom hierarchies or sync with vCenter Server

– Setup Cost models, base rates, billing policies and other fixed or operational costs

– Setup cost template and rate factors

You can also view excellent webcasts presented by Lawrence Koh on vBrownbag sessions here.

Quick Takeaways while installing NetApp FAS Series SRA 2.0.1 with SRM 5.0.1

I am currently deploying an SRM Solution with the following config and there were so many errors to watch out for:

Protected Site:

vCenter 5.0.0 build 455964

SRM: 5.0.1 build 633117

NetApp FAS Controller: FAS3270 (Active Active Controllers)

NetApp FAS/V Series SRA: 2.0.1

Recovery Site:

vCenter 5.0.0 build 455964

SRM: 5.0.1 build 633117

NetApp FAS Controller: FAS3240 (Active Active Controllers)

NetApp FAS/V Series SRA: 2.0.1

So first and foremost thing, do not install SRA 2.0.1, the moment you try and pair the Array managers, you will get this below error:

“Internal error: std::exception ‘class Dr::Xml::XmlValidateException’ “Element ‘SourceDevices’ is not valid for content model: ‘(SourceDevice,)'”

This is because of a bug confirmed by NetApp, Bug ID: 642115

http://support.netapp.com/NOW/cgi-bin/bol?Type=Detail&Display=642115 (will require a NOW login)

So, the workaround is to either include the volumes manually that you want your Array manager to detect or simply fall back to NetApp SRA 2.0.0 (which was the quickest option for me looking at the amount of volumes we had on both the controllers)

Now then, I was  unable to add 2nd active controller (Error: SRA command ‘discoverArrays’ failed.), although IP was pingable, was able to ssh into the filers etc…so how did I resolve it? by running these simple commands on all the controllers:

options httpd.admin.enable on

options httpd.enable on

options httpd.admin.ssl.enable off

A quick reminder on NetApp Volume design:

You need to ensure that your VMDK’s or RDM’s does not span across controllers (either at Prod site or at the recovery site), this is not a supported configuration by NetApp

I shall keep updating this post as I move forward with my configuration 🙂

vCenter Operations Manager – Complete installation guide with screenshots

The vCenter Operations Manager and Infrastructure Navigator vApp is easy to install but vCenter Configuration Manager requires bit of an effort.

I have attached the document which lists the step by step instructions with screenshots. Some pre reqs to be done before installing vCenter Configuration Manager:

  • In a single Tier setup all the 3 components i.e. DB, App and Web are installed on the same Server either physical or Virtual
  • In a Two-Tier split setup, install the SQL and VCM Database on one machine and IIS, SSRS, VCM Collector, VCM Web Console and VCM Reporting role on a separate machine
  • Database must be installed with server collation  SQL_Latin1_General_CP1_CI_AS, if not uninstall and reinstall SQL
  • The default encoding for collector and database server must be English (United States) 
  • I have listed the required service account and their permissions in the attached word file
  • While in Two Tier setup, install the SQL (ensure mixed mode authentication is selected, otherwise VCM installer wont detect the SQL instance) and VCM DB Component first, then install IIS, SQL Reporting Service and configure them as mentioned in the VMware Installation guide
  • Only use NetBios name while installing VCM otherwise the installation will fail at 90% which is frustrating
  • The VCM Collector component installation will take a hell lot of time while installing the 16th and 17th package, so please be patient

Please feel free to add any critical information that I may be missing

vCOM’s Installation_Screenshots

vCloud Director Installation Pre-requisites

This blog post is not about vCloud Director installation rather working out the pre-requisites which are hard to work on if you are not from a Linux background:

1) Installing Red Hat 5.x on a Virtual Machine must be straight forward, but if you still need a step by step instructions with screenshots then follow the link here

2) Installation of VMware Tools is required if you provide vmxnet3 NICs to your VM.

Click VM in the virtual machine menu, then click Guest > Install/Upgrade VMware Tools and click OK

mkdir /mnt/cdrom

mount /dev/cdrom /mnt/cdrom

rpm -ivh /mnt/cdrom/VMwareTools-<version>.rpm

Where <version> is the version of VMware Tools you want to install

To check the existing version of VMware Tools, use the “cd” command to go to /mnt/cdrom/ directory and list the contents of the directory with the command # ls.


umount /mnt/cdrom

You can even follow VMware’s KB article here and here

3) Changing IP addresses on eth0 and eth1 interfaces is easy if you have used the VI editor before, or just follow this link here

4) Database creation is again easy as it is mentioned in the VMware documentation, run those scripts as is, just by changing some values i.e. the path to your Database, DB name and login credentials

5) Create SSL certificates, this can be a little tricky if you have not done this before:

There are 2 ways to do it:

a) Either create them on a separate machine and copy them over, which is a little tricky and tedious

b) use the keytool provided by vCloud Director installation.

The question is, How do I actually use the keytool before even installing vCloud director?

Well its simple! you need to install (by running the installation command provided in the VMware documentation) the vCloud Director and “DO NOT CONFIGURE IT”

Once the installation is done, browse to /opt/vmware/cloud-director/jre/bin/ and type these commands:

keytool -keystore certificates.ks -storetype JCEKS -storepass password -genkey -keyalg RSA -alias http -keypass [password]


keytool -keystore certificates.ks -storetype JCEKS -storepass password -genkey -keyalg RSA -alias consoleproxy -keypass [password]

Remember, you will need to enter some information manually, they are self explanatory and should be easy to enter

once you have generated the certificates, the run the vCloud Configuration i.e. /opt/vmware/vcloud-director/bin/configure and follow the onscreen instructions

6) Accessing the vCloud Director Portal from the vCloud Director server requires you to install the Flash Plug in. Just follow these steps below to install Flash plugin on the Linux box:

download “flash-plugin-” (check for latest release on the Adobe website) and move the rpm package onto any folder on the vcloud director server

Browse to the location onto which you moved the rpm package and run  “rpm -ivh flash-plugin-” and wait for it to install”

Please let me know if anything is missing or can be reworded with proper steps and bullet points.

Windows 2008 R2 64 bit V2V conversion fails with “Unable to determine Guest Operating System” error

Windows 2008 R2 64 bit V2V conversion fails with the below error:










The V2V was across the VC with both the source and destination (vCenter and ESXi) being on 4.0 U2 and VMware convertor running on 4.0.1 Build -161434

Resolution: Just change the guest OS type (Right Click the VM> Edit Settings> Options tab)  to anything below 2008 R2 on the source machine and the V2V should go through fine

Thanks to: http://wiert.me/2010/06/10/vmware-converter-4-01-unable-to-obtain-hardware-information-or-unable-to-determine-guest-operating-system/