The enStratus Blog

Cloud Computing and the Business Model

Everyday we talk with companies that are evaluating cloud computing capabilities to reduce costs, be more flexible, leverage new technologies or deliver scale. More and more enterprises are looking to cloud computing and are starting proof-of-concepts and migrating real applications to leverage these benefits.

There is a deeper question forming about how business models are impacted by cloud computing and vice-a-versa. Most of the cloud talk to date has been focused around the technical challenges, opportunities and questions. However, not enough has been focused around the business model transformation of the cloud.  

For several years now, software vendors have grappled with business model questions and monetization issues as they transition to a SaaS delivery approach. Now with the barrage of cloud services, Systems Integrators, Hardware Vendors and other types of technology firms also have the same types of challenges.

These points are highlighted nicely in the new Accenture Report titled Where the Cloud Meets Reality: Operationally Enabling the Growth of New Business Models. We recently discussed this new Accenture Report with one of its authors, Tim Jellison, and we agree with the business model topics identified in the report.

The Accenture Report identifies ten different business models (ie Licensed Software, PaaS, Appliance Model) and discusses how organizations should address and plan for the operational model changes and how to get started.

There is no question that the XaaS market is growing rapidly.  The Accenture Report highlights a Gartner study that shows by 2014, enterprises will speind close to $35B annually on SaaS, PaaSm IaaS and other XaaS models combined.  That is up over 300% over 2010.

Explosive growth and disruptive technologies brings more than revenue opportunities to companies like enStratus. It brings questions that finance, technology, operations, legal, marketing and sales leaders need to address.

This is probably the start of the business model conversation, but it is a conversation that has to happen on many fronts.

Multi-currency Financial Controls

enStratus financial controls enable you to track your cloud spending across multiple accounts in multiple clouds down to the resource level and associate cloud resources with distinct cloud budgets. How do you handle a multi-cloud environment, however, that includes cloud providers who charge you in different currencies? Or what if you are using a single cloud provider who charges you in a currency different from your national currency? Today, enStratus announces multi-currency financial controls to enable you to maintain budgets in your company accounting currency across clouds billing in any number of different currencies.

As we continue to gain customers across the globe and expand our cloud support to clouds that bill in currencies other than US dollars, we made sure to extend our financial controls to operate in this environment. When configuring your account, you can now specify the company accounting currency to be whatever currency you commonly use internally. You establish all of your soft and hard budget quotas in your internal currency, and those are the values your team works with in enStratus. enStratus also displays reports using your accounting currency.

Behind the scenes, your cloud providers may be billing you in any number of different currencies. These currencies may not match at all to your internal accounting currency. enStratus will track your spending against budgets by translating the prices from the cloud provider currency into your accounting currency. We use either exchange rates updated daily or exchange rates you set for your organization. If your spending in Australian dollars in one cloud combined with your American dollars costs in another cloud exceed your budget defined in Euros, enStratus will alert you and/or prevent further resource allocation against that budget.

enStratus and GoGrid

 enStratus already manages more public providers and private cloud technologies than any other cloud infrastructure management platform. Today, enStratus announced support for one of the pioneer cloud providers in the Infrastructure as a Service (IaaS) cloud space, GoGrid. This match teams the first provider to support Microsoft Windows in the cloud with the first infrastructure management provider to support Windows in the cloud.

As you would expect, enStratus enables you to provision and de-provision resources inside a GoGrid account. That's just the start. You also get all of the governance features you need for managing any cloud environment in accordance with typical enterprise IT requirements. These capabilities include the ability to establish different billing budgets for resources within the same GoGrid account, custom user role mappings with fine-grained permissions, LDAP and ActiveDirectory integration, management of shell and RDP accounts on your virtual machines in the cloud, and many other enStratus features.

Over the next month, we will also grow our support of GoGrid to support pre-bundled machine images with the enStratus agent pre-installed and infrastructure automation, including auto-scaling, auto-recovery, backup automation, and more. This new support will include the ability to leveraging cloud bursting from a private cloud into a GoGrid public cloud.

Every CFO Needs To Have “The Cloud Conversation”

Dear CFO,

Your IT infrastructure is at the beginning of an important transition and if you haven’t had “The Cloud Conversation” with your CIO or CTO, you should.

The cloud you need to talk about is the infrastructure cloud - the kind of cloud Amazon has pioneered with its Amazon Web Services (AWS) offering. Amazon and friends are changing the economics of IT infrastructure thru the application of scale and specialized expertise in much the same way that Amazon.com changed the rules for the smaller book store.

There is a whole separate conversation to be had concerning the agility, flexibility and transparency that comes with a cloud infrastructure - but I am going to try and focus here on the financial and economic advantages and prepare you for “the conversation”.

The big picture is that there are three economic legs to the cloud infrastructure stool: Scale, Specialization and Utilization.

Scale

The scale advantage is simply that Amazon, Microsoft, Google and others are investing billions of dollars in building out state-of-the-art data centers and then renting usage back to customers by machine hour - the price point has already come down as low as a few pennies per hour -  so you get a small piece of their buying power and because you only pay for actual usage, by the hour, no CapEx is required.

So the first response from your CIO might be that your current hardware costs equate to  similar or even fewer pennies per hour when costs are amortized over 3 years, even after cost of capital is included - and they are probably right - the cloud providers are adding a margin back into the price they charge you!

Specialization

BUT, and here’s the second leg of the stool - servers and storage are probably only about 50% of your IT delivery costs (and I’m excluding the cost of the DC build out and the cost of people). Power (and the equipment required to keep your hardware cool) is the other major cost component and it has been growing over recent years. In older, less efficient  facilities, this can be a larger cost that the actual servers and storage. If you are using a colocation facility, you are probably very familiar with the costs I’m talking about here.

The infrastructure cloud providers are at the forefront of innovation to increase energy efficiency and reduce this cost - they are quite probably between 50% and 100% more efficient that your current environment today, and investing substantial sums to get better.. 

The bottom line is that after combining hardware and power/cooling costs, you are likely already paying more than the price available from an infrastructure cloud. And we still haven’t talked about the most important advantage of a cloud infrastructure .....

Utilization

The third leg of the argument, and the one that uncovers significant advantages for many, is a result of the hourly usage pricing model. Most data centers operate at 10%-20% utilization, some as low as 5%. This means that at least 80%-90% of your infrastructure is sitting idle for most of the time - adding no value, just costing you money. This extra capacity is there for those few weeks, days or even hours when you hit peak demand - end of year sales, end of month processing, major promotions etc.

Simply by moving to a usage-based payment model in a cloud that scales resources up and down as required, you will eliminate most of the cost of idle capacity.  The implications for your organization will depend on your current utilization and the demand variability of the application portfolio, but it is not at all uncommon to see a 50% or greater cost saving when applications are moved from a traditional data center or hosting provider to a cloud infrastructure.

At this point in the conversation, if not before, you might hear from your technical team that the public cloud infrastructure is too risky and that there are security concerns. 

But the main points I would suggest are that (a) the object here is to start down a path with a pilot project, not start with the crown jewels, (b) there is already a ton of money and many companies focused on addressing these issues to a level well beyond their resolution in most internal data centers today, and (c) there are many very large and conservative organizations who have found enough here to start down their journey.

Can Private Clouds Get Me There?

Another response you should be prepared for is that the team is already well ahead of you and evaluating a private cloud strategy. This is great news.

But private clouds have a different economic model - they involve you buying the infrastructure and delivering on demand computing resources across your business. So the business advantages of responsiveness, agility, and transparency should be the same as for a public cloud like AWS, and your team will have greater control of the underlying physical infrastructure. But you may not get the economic advantages of scale, specialization and utilization we have been talking about here.

The traditional infrastructure vendors are very keen on this private cloud model - for the obvious reason that they can still sell you all the same kit. So be aware! This could be a great solution for your organization, especially if you are very large and can consolidate IT infrastructure resources across divisions or businesses. It will not automatically deliver the cost advantages spelled out above. My recommendation would be to push for a parallel project to try out the public cloud and provide context and comparison for private projects.

So, my Dear CFO, I hope this helps equip you for the conversation. Cloud infrastructure is not a silver bullet. Like all new things it has its challenges - but it is important and will change your IT cost model and that of your competitors. Now is the time to ensure that your IT team is engaged and gathering experience and perspective to build out a plan that is right for your organization.

It's Official: enStratus Rolls Out vCloud Express Support

 A few months ago, we quietly mentioned in this blog how enStratus uses Dasein Cloudas a cloud abstraction layer and, thanks to Dasein Cloud + jclouds, that enStratus would support Terremark's vCloud Express through Dasein Cloud + jclouds + vCloud. Today, we formally announced vCloud Express support with Terremark as our launch partner for this support.  

enStratus supports all of the core features of Terremark vCloud Express, including the provisioning and de-provisioning of virtual machines and allocation IP addresses with the ability to setup IP forwarding rules. 

With enStratus, however, you expect more than simply being able to access the core Terremark functionality. enStratus provides all of the governance elements for the Terremark cloud that you expect, including:

• Cross-account user management
• Role-based access controls for managing your Terremark infrastructure
• Financial controls to enable you to track spending
• The ability to manage Unix shell and Windows RDP users from the enStratus console
• The ability to setup modjk, haproxy, and Zeus load balancers.

In our Efreet release scheduled for early June, we will add even more capabilities to our Terremark support:

• Cross-cloud clustering with auto-scaling and auto-recovery (burst from a private cloud into Terremark or balance load equally across two clouds)
• Automated disaster recovery execution into a backup cloud
• Auto-scaling and auto-recovery with automatic load balancer configuration in Terremark
• Mix Terremark compute resources with cloud storage from another cloud to create a single logic cloud
• Support for Terremark volume configuration

Finally, we'd like to take a moment to thank everyone at VMware, Terremark, and jclouds who have all been very suppportive of bringing this effort to market.

Update Dryad

We will soon be rolling out Update Dryad which is a major system update with significant new functionality for enStratus. Major enStratus updates all center around a specific theme. The theme for Dryad was "user management".

One of the strengths of enStratus has been the ease of user management in the platform. It's one of the oldest elements of the system, and we used Update Dryad to significantly move our user management forward. Dryad introduces:

• User groups (users can be in one or more groups)
• Roles, with each group assigned a specific role on an account-by-account basis
• Comprehensive access rights management for roles

Once the upgrade is complete, you will have four user groups attached to four roles of the same name in each account. These groups and roles match up to the old Admin/Server Manager/Configurator/Reports permissions.

 

You can immediately begin customizing the system to suit your needs.

First, you define roles to encapsulate a set of access permissions. An access permission defines an action on a resource type with an optional qualifier. Qualifiers limit the scope of an action on a resource. For example, you can define a permission like:

SERVER/Terminate/MINE

That permission means that anyone tied to that role can terminate only servers they launched. You can also qualify an action based on group membership or billing code.

Why the separation between roles and groups? Within the next couple of months, we will introduce ActiveDirectory and LDAP integration. User membership in groups will be governed by your AD or LDAP management policies. Access to enStratus resources for customers leveraging AD or LDAP integration will thus be governed by tying an AD/LDAP group to an enStratus role. Users and groups will be "read-only" in enStratus for such customers.

A few weeks ago, we introduced our plans for the finance view into the system. Our next major update, Update Efreet, is focused entirely on the finance view of the system. We set the foundation for the finance view, however, in this release:

• You can define a set of billing codes and associate resources with those codes for tracking cloud usage/billing
• You can view your enStratus invoices from inside the console

To help with compliance, we've exposed more of our underlying audit capabilities in the form of Firewall change log reporting. From the reports console, you can view all firewall rule changes that have been made to the firewalls in your account, including changes made outside of enStratus. It also provides for a CSV download of your firewall change log. We also now send an alert any time a firewall rule changes.

Other changes in this release include:

• Support for AWS load balancers
• Greater ability to control AWS auto-scaling rules from the UI
• Beta support for SAML 2.0 authentication
• Configurable named networks and services to make firewall management easier
• Support for Terremark vCloud Express
• Greater ability to control alert preferences

Managing Cloud Budgets with enStratus

About two weeks ago, we discussed how to use enStratus labels to help prevent VM sprawl. As an organization adopts cloud computing, however, VM sprawl is just one of the new challenges cloud computing presents. Over the next quarter, enStratus will be introducing a number of features to return control over finances and budgets back to the finance department.

Companies often adopt cloud computing to reduce procurement windows and eliminate capital expenditures. These benefits come with the side-effect of reducing the ability of the finance department to prevent IT from over-spending or managing who is spending approved funds on what resources. enStratus provides that control.

Billing Codes

In our next release in January/February, we will introduce the ability to assign billing codes to users and define budgets against those billing codes. When users launch virtual machines, enStratus will track your enStratus and cloud costs against a billing code assigned to that user. As usage for a given month appears to put the budget for a billing code at risk, enStratus will send members of the finance department an alert enabling them to take action.

We will also be tying billing codes into the custom role definitions we are also rolling out in the next release. This feature will enable you to define permissions such as "able to shut down servers associated with my billing codes".

Unified Reporting

This next release will provide monthly reports that will identify all costs associated with your cloud infrastructure across all clouds. The report will break down by billing code and include estimated cloud provider costs as well as enStratus costs. You will be able to download these costs and important them into your internal financial control systems.

Shared Resource Accounting

Another challenge in cloud computing is the management of budgets allocated to shared infrastructure. When you deploy a set of servers with a cloud provider using the enStratus cluster manager, you can assign different billing codes to the services running in that cluster and enStratus will split the infrastructure costs across the associated billing codes.

Bringing It Together

The net impact of these enhancements is that your finance department will maintain control over the financial impacts of IT procurement without creating the unwieldy procurement processes that cloud computing is supposed to help you escape from. Finance will be able to create budgets for different groups and monitor the spending across those budgets. IT will still be able to procure within the constraints of those budgets and know in advance if those budgets are at risk.

Using Labels to Prevent VM Sprawl

Last month's update introduced the ability to label servers with a special color. enStratus assigns no meaning to this color. Instead, you can label a server with a color and place any meaning you like on that color.  

From our perspective, one of the most compelling uses of labels is to help control VM sprawl.

VM sprawl happens when people indiscriminately start up virtual machines in the cloud but hesitate to shut them down. Basically, it's much easier to start virtual machines in the cloud than it is to shut them down. Starting a VM is never something destructive; shutting one down is. People consequently hesitate to terminate a cloud instance unless they know for certain that its termination won't have any negative consequences.

enStratus color labels can help you color code your cloud servers so that it's easier to tell what servers can be shut down.

For example, you might establish a process that any unlabeled servers in the cloud can be terminated by any administrator 48 hours after launch. If you want a server to live longer than 48 hours without someone else shutting it down, you must assign it a label. Some labels may represent production systems, while others may represent test or R&D. These colors not only make it easier to kill off servers that were meant for temporary lives, it also makes the process of reviewing what survives simpler.

In the December update, we will expand the presence of labels to machine images and provide the ability to set a label when you launch an instance.