I’ve been hearing more about customers embracing ESXi, and even more that some are beginning to deploy it at scale in production environments. I want to share a few brief notes of concern that I have validated with other large customers and with some hardware manufacturers.

The VMware kernel part of ESXi seems to be rock solid, same as their other core hypervisor stuff.  So the concerns outlined below are not meant to say that VMware’s ESXi is not ready for primetime production use, but rather as suggestions to make sure it’s right for you.

Several customers, and two of the big three hardware vendors, are raising a flag and saying that customers should be aware of what hardware failure and performance alerts they can’t receive on ESXi hosts.  Customers can minimize the impact of these concerns with the following recommendations:

• Customers should perform extensive testing in their environment with ESX 3.5 beside ESXi and compare the alerts each generate from such failures as the removal of a power supply, drive, array controllers and memory errors.
• Customers should only run non-critical loads (dev/test/etc) in ESXi for a measured period of time to fully understand its performance and monitoring characteristics in their specific environment.
• Customers should establish a conversation with another customer that has been down the path of trying ESXi in their environment.  These peer-to-peer conversations really do add a lot of real world experiences to the conversation.
• Customers should take a close look at all of their infrastructure tools in the new environment: backups, replication, security, audit, monitoring, asset management, capacity planning, provisioning, etc.  Many might not function since their agents are longer able to exist in the console OS.  Others could be significantly impacted in performance due to alternate means of execution.

The bottom line is that I’m not suggesting that ESXi is not a great fit for customers in their environments, but rather that most customers experience some significant pain while they learn how to manage ESXi in their unique environment.  Vizioncore can be a trusted advisor by merely pointing out some of the pain points, such as those above that have been experienced by others; however we are not typically able to assist further than referring them to a partner to help with the VMware side of the execution with products other than our own.

I am very interested in your experiences around this topic and would welcome any corrections, or alternate views that you have.

Sincerely,

Brad Wagner
Director of Product Strategy
Vizioncore

Last month I attended VMware’s Partner Exchange in Orlando, FL where I had the pleasure of meeting with many VMware VARs, VACs, Technology Alliance Partners, Service/Hosting Providers and other virtualization professionals.  I want to take a few moments and share some of the common themes from our conversations and use them as a starting point for this blog, and for future blog discussions.

First (and most evident) was the understanding and focus on the tremendous market opportunity that surrounds the virtualization industry.  Every keynote, breakout session and personal conversation reiterated the multi-billion dollar market for solutions that address key needs in virtualized environments.  Furthermore, the areas of concentration for these opportunities continue to multiply as businesses realize the value of virtualization and embrace new technologies to help them optimize their environments.

However, there are common concerns about the rate of adoption in the overall marketplace.  Many of these concerns are repeated by the commercial and enterprise customers I consult with and have been a road block to their perceived “utopian” path to a 100% virtual infrastructure.  So what are the concerns?  Each of these warrant their own blog discussion which I’ll address in future articles:

• Expertise, Attrition, Complexity
  Finding and retaining employees with true expertise in SAN, Network, Security, Hypervisor, OS and Application troubleshooting and optimization is a challenge.  Also take into consideration that when you lose an employee at this level your also losing their inside knowledge of your unique infrastructure, let’s just hope they left behind proper documentation.

• Management Tools & Root Cause Determination
  There is an immense opportunity for comprehensive management tools and solutions to quickly pinpoint the true root cause of issues encountered across all layers in a virtual environment.

• Costs
  While virtualization often provides an almost immediate ROI, businesses that mature in their virtual endeavors will face more challenges resulting in additional costs.  If you only look at the license and initial hardware price you’ll be sorely surprised when you see other costs for management, process changes and growth.

• Barriers to Widespread Implementation
  Businesses can be initially encouraged by the simplicity of implementation but will encounter many growing pains as they expand their virtual environments.  A common request I receive from customers is for clearer guidance during implementation to avoid common pitfalls.

Another interesting topic from the conference I want to callout is for solution providers to have a better understanding of the needs for SMB customers vs. Enterprise customers.  The features and analysis requirements in each of these segments is very different and requires more attention to their specific needs in order to be successful when selling to them.

Overall this springs’ Partner Exchange was well received by all.  I was very encouraged to see partners maturing in their understanding of virtualization and the many challenges and opportunities it provides.  I look forward to providing you with more blogs with a greater focus on the areas highlighted above.  In the meantime I invite you to share your comments, questions or other concerns you had from the conference.

Sincerely,

Brad Wagner