Translate this website:
Search this website:


CitrixDesktop VirtualizationHyper-VI/O VirtualizationNetwork VirtualizationRedHatServer VirtualizationStorage VirtualizationVMware

Could Virtualization Run Aground?

asks Jan Ursi EMEA Infoblox.

 

Date: 21 Nov 2011

Most IT revolutions have their tipping point, where the market expands beyond brave (or reckless) early adopters, through mainstream acceptance to suddenly becoming a universal ”must have”.

Virtualization has certainly passed beyond the first stage and, as a concept, is widely accepted – with nearly 40% of x86 workloads already running in virtual machines at the start of this year, and expected to grow to over 75% by 2015 according to Gartner (March 2011 data).

But the tipping point to mass acceptance would mean spreading from the deep waters of Fortune 500 companies and reaching into the broad shallows of the SME market. Are they ready for that?

Judging by the disturbing results of a recent survey by Infoblox, the answer is a definite ”no”. The good ship Virtualization is about to run aground…


Digital systems built on paper foundations?
Although a recent survey by Infoblox set out to track IPv6 readiness across the EMEA – A key question asked was ”How does your organisation currently track IP addresses?”. The answers to this question are very significant in the context of intended virtualization.
40% were still relying on manual processes – at best keeping track on spreadsheets or even relying on paper records.
Although over half have migrated to automated software, the fact is that in most cases this was simply provided by bundled solutions such as Windows Active Directory. Though not recorded in the survey results, it does match analyst’s estimates of the overwhelming reliance on either manual processes or the simplest tools for managing network administration.

These simple tools are cheap and ubiquitous, but can usually only handle multiple domains and subnets by providing coordination outside of those systems – and such coordination is itself often based on manually maintained spreadsheets. Furthermore, in the many organizations that operate both Linux and Windows, some form of manual coordination between the two is also required.
Besides the labor cost and the opportunity for errors, such solutions pose problems for industries under compliance regulations. Neither BIND nor AD nor manual spreadsheets have built-in history reporting and audit trails. For compliance reporting, these are needed on a regular basis, and require even more tiresome manual labor.
The demands of virtualization
True, virtualization is already widespread – but for the majority it goes no further than simply, ”phase one” installations that enable multiple instances of applications to be tied to a single physical server to reduce space, and power and cooling costs.

If enterprises try to scale beyond phase one to support dynamic workloads and private cloud infrastructures, they will be faced with unprecedented levels of change and complexity. In the virtual data center, administrators must be able to re-provision processing power at a stroke. If this change is not highly automated, the many steps involved in re-provisioning the network infrastructure and its core services such as IP assignment, DNS and DHCP, will demand considerable time and manpower. These changes include: firewall, VLAN, QoS, and policy settings, and other changes to both physical and virtual network elements to support the virtual data center.

While this rate of change and complexity is accelerating, few IT departments are expanding to keep pace. As Zeus Kerravala, Senior Vice President and Distinguished Research fellow at Yankee Group explained:

“Throughout a virtual machine lifecycle, there are multiple events that require visibility and change; most server-centric functions for managing this are built into the hypervisors, but the flip side of the coin – causing tremendous complexity as enterprises try to scale their virtual deployments – are the associated network changes that are typically performed manually. Automated technology is the only way to overcome this hurdle.”

Nor is it just a question of managing change, for there is also the problem of keeping subsequent track of those changes, as Daniel Boyd, IT administrator at Berry College points out:

“Virtualization has been touted as the best thing since sliced bread in terms of cost savings and flexibility. We have realized part of that promise. But, there have been some visibility and management issues. It’s still daunting to know where everything is and it would be helpful to have a single application that shows us where all the resources are and which virtual machine is on what host with all the information in a single view, instead of having to check five different management applications to find the information I need.”

Berry College’s experience is far from unique, according to industry analyst Jim Frey of Enterprise Management Associates, who adds:

“There is little if any hope for manual processes to keep pace with the rate of change introduced by server virtualization and cloud services – the only reasonable answer is automation. In this case, network managers could benefit greatly from tighter automation and control around IP address management as an essential aspect of maintaining a highly functional, highly performing network.”


Virtualization on the rocks?
The upshot is that virtualization will never get beyond phase one until network managers stop relying on spreadsheets, manual processes and basic bundled tools for tracking IP addresses and managing core services such as DNS and DHCP. They must migrate to an automated solution that can scale in a virtual environment where virtual machines are dynamically provisioned, migrated and shut down.

Yet, according to our survey, 40% of organizations have not even begun to address this problem, and those that have are still predominantly relying on the most basic software tools. Even without the added pressures of virtualization, the analysts’ statistics are disturbing:

· Two thirds of all system performance issues are linked to network change, according to Gartner and IDC
· Over 70% of IT staff’s time is spent on unplanned work according to a study by Metzler & Associates
· Up to 80% of all IT resources are consumed simply to maintain the status quo, according to IDC

Without extensive automation, this wasteful situation will continue, and virtualization will remain a niche technology for all but a few of the largest organisations. Because it will be enjoyed only by those that have already invested in flexible, automated systems that deliver the necessary visibility, change management and compliance capabilities in both the physical and virtual environments.

What is the secret of their success? It is no secret that today’s sophisticated and yet simple to deploy solutions for network change and configuration management combine powerful automation together with clear visibility into the health, policy and compliance of the network. They can collect and analyze network infrastructure configurations, identify policy violations, show the impact of change on network health and remediate issues. In short, they enable the enterprise to automate core network infrastructure management and support highly dynamic networks, applications and initiatives – such as virtualization and cloud computing.

If that was not enough, they also provide immediate ROI in terms of less reliance on specialist IT staff, many fewer repetitive manual tasks – saving time and reducing the risk of human error – and reducing operational and logistical delays that hinder business agility.

Conclusion
Virtualization and cloud computing are technologies that promise huge benefits for the whole business community, but the vast majority of enterprises do not yet have the fundamental network capabilities needed to reap those benefits.
Before even looking at a server virtualization or cloud solution, the first question should be whether the network is ready to support virtualization. Without network infrastructure automation the answer will be a clear ”no”.
This automation does not demand a massive and disruptive change to the network. Today’s solutions can simply be plugged into a network – or loaded as software into a virtual environment – for control from a convenient console. Nor does automation demand massive investment, as entry-level units are available for smaller branch offices, right through to top of the range systems for the largest datacenters.
If it were only a question of enabling migration to tomorrow’s virtual environment, it would be a strong enough argument. However, network change automation also addresses the dominant problem dogging business networks today – 80% of all system outages can be attributed to the human errors that arise when changes are made to the network.
Automating core IP network services such as IP adress management, DNS and DHCP, and integrating it with automated change on all network devices such as switches, routers and firewalls – that’s the only way to prepare today’s networks for the demands of the future.

ShareThis

« Previous article

Next article »

Tags: Server Virtualization, Storage Virtualization, I/O Virtualization, Network Virtualization, Desktop Virtualization

Related News

16 Apr 2014 | Server Virtualization

16 Apr 2014 | Server Virtualization

16 Apr 2014 | Server Virtualization

15 Apr 2014 | Server Virtualization

Read more News »
Related Web Exclusives

14 Apr 2014 | Server Virtualization

7 Apr 2014 | Server Virtualization

17 Mar 2014 | Server Virtualization

24 Feb 2014 | Server Virtualization

Read more Web Exclusives»

Recruitment

Latest IT jobs from leading companies.

 

Click here for full listings»

Advertisement