Cisco Blogs


Cisco Blog > Data Center and Cloud

Good post on cloud journeys: crawl, walk, run

Cloud is a journey. This post discusses our approach to crawl, walk and run.

A cloud architecture has multiple facets and requirements, a key part of which is the need for cloud orchestration and provisioning, coupled with a self-service end user portal.  Let’s call this “Cloud Automation” for now.  If you are designing and/or building a cloud, then, part of your work will be to deliver a cloud automation solution to deliver on that promise.  How do you plan to go about that?  One approach is to define your extensive list of requirements, based upon your business needs and current capabilities, and go about building out that solution.

Another approach is what I’ll call “Crawl Walk Run”.  The incremental approach.

Post is here.

Cloud is a change to the operational model: a change in behavior, accounting, process and people. You can’t do it overnight. Trying to deliver every service doesn’t work.

It’s very important to set a roadmap of where you want go with your cloud services so you don’t get stuck in the VM Azores — this is where all the focus is on VM provisioning and then you deploy technology that does that. And only that.

You need that roadmap of services and a technology platform that supports your vision. Even if all you first is crawl.

 

Tags: , , , , , , ,

The Road to PaaS. What’s Post-IaaS – Network thoughts

Recently, I wrote an article on PaaS for IT BusinessEdge entitled the road PaaS, understanding your post IaaS options.  Here’s an excerpt.

The Road to PaaS

PaaS is an enticing proposition that has generated a lot of market buzz.

But PaaS forces tradeoffs and it shouldn’t be seen as a one-size-fits-all proposition.

To understand, I like to draw the distinction between what I call “Silicon Valley PaaS” and “Enterprise PaaS.” The majority of the discussion in the market today revolves around the Silicon Valley PaaS pattern, which is a truly abstracted “black box” approach to software platforms.

This form of PaaS exposes a set of standardized services to which you write your applications, completely sheltering developers from the underlying complexity below the PaaS abstraction.

It makes a lot of sense for brand-apps built with modern frameworks like Python and Ruby in greenfield development environments that are highly standardized.

The basic premise of the post is that PaaS for an enterprise is VERY different from PaaS for a Silicon Valley start up. And nowhere is it more  different than in the network requirements.

The PaaS customer is a developer who will code an application, use the underlying services offered by the PaaS stack, such a database, storage, queueing, etc.  The developer deploys the code, selects a few options and code is live.

So what’s going on with the network? Well, the PaaS layer will need to auto-scale, fail-over and deliver performance at some level. It may need it’s own domain as well. That PaaS layer will need to talk to underlying network services such as firewalls, switches, etc.  That PaaS really needs access to infrastructure models that deliver network containers to whatever PaaS abstraction the PaaS layer has.

Hard enough to do when all the containers are the same, as it would be in a Silicon Valley PaaS offering.

It doesn’t work with the existing enterprise platforms.  This is a big opportunity for innovation

Tags: , , , , , , ,

Part 5: A service catalog will help Vmware admins get ready for cloud computing, public or private.

Part 5 of the series “10 Things VMware Server Admins Should Know About Self-Service Catalogs and Lifecycle Management” that I’ll be publishing over the next couple of weeks.

5. A service catalog will help VMware admins get ready for cloud computing, public or private.

When I first came up with the concept of a service catalog to drive fulfillment process back in 1999 (Yep. 10 years ago. Time flies when you are having fun.) it was obvious that internal shared services like IT needed to emulate the likes of Amazon.

Well, here we are in 2009 and the wheel of time has brought us back to the same place.  Now it’s the data center that is being disrupted rather than end user services.   Customers are beginning to ask:  Why can’t you be more like Amazon EC2? Why can’t you provision fast, at guaranteed cost?

Let’s look at how Amazon EC2 uses the concept of a service catalog and lifecycle management to deliver cloud computing in consumer-like experience.

There’s a lot of talk about the technical aspects of cloud computing, and little the customer side:  Amazon communicates with its customers through a service catalog and lifecycle system.  The brochure part of the catalog is found here.  (I wrote this in more detail in my post: Amazon has written your technical services catalog).

To see the full functionality of this service catalog in action, I broke it down into Structure, Benefits, Pricing and Actionable for simplicity.

Structure

The whole structure looks like this:

It covers what it does, what benefits (hightlights), details, major options and pricing! Then what I call the fine print (aka SLA’s).

Benefits

It doesn’t skimp on benefits.  In fact, benefits and outcomes are front and center. We can do the same with with our virtualization offerings.

They tout their unique differentiators are variable (elastic) cost, while re-assuring that you have complete control, flexibility and of course, it’s inexpensive. In fact, if you read that section, it draws a comparison against an internal data center!  And it gets to heart of what customers don’t like about IT costs; highly fixed, over-bought, hard to plan for, etc.

It also covers the OS, database software and middleware choices. This is an example of going beyond the server.

What are your benefits? What are your unique differentiators?

Pricing

Next, the catalog outlines the main packages: Standard and High CPU.  Two choices, and then some three sub-choicess.

There’s a lot more description, links to explanation, FAQs, etc.  It’s the way they standardize these formerly complicated configurations that is a useful take away.

Pricing follows and there three aspects to highlight.  First, it’s completely and easily understandable as a unit of measure. They use per hour.

Standard Instances Linux/UNIX Windows
Small (Default) $0.10 per hour $0.125 per hour
Large $0.40 per hour $0.50 per hour
Extra Large $0.80 per hour $1.00 per hour

Think of all the complexity of running a datacenter: people, machines and facilities, etc. Amazon gets it down to controllable unit of of measure, hours. As a customer, I can choose to consume and hour or not.  That’s a level of control that’s appealing to me.  Is this the right unit of measure for every customer? No. It will depends on your customer and the benefit they want to buy. (More in future postings).

Second, they include all the pricing units for network, storage and servers.  Your complete datacenter (almost) configuration.

Third, some charges like data transfer charges are harder to map to controllable costs, so Amazon provides a pricing calculator to help translate these costs into the potential bill. And they provide sample configurations and estimates.

Except for chargeback, which you are doing or not, every leson is directly applicable to how we present virtual environments.

How does the catalog play a role? In two ways, it establishes the standards which enable self-service and then uses those to meter and report to your account what your consumed.

Actionable!

Finally, this catalog is NOT STATIC. It’s completely actionable.   If you have an account and log in, Amazon provides:

  • Self-service ordering, configuration and deployment. This request management against known, vetted standards is core to making cloud computing work. Think if Amazon had to go back and forth for weeks with a user about their configurations?
  • Account management functions. The customer can perform a variety of actions on their own to manage the lifecycle of virtual instance.
  • Consumption management and billing.  The customer gets clear, hourly consumption metrics.

In other words, Amazon delivers a very complete service catalog tool set to enable cloud computing. I like that they have brought the ease of their regular catalog to a more complex environment. And ease wins.

Amazon has redefined the expectations and pricing for data center services. Make no mistake, they are your competitors. Now the challenge is to respond with your own service catalog and differentiated service definitions.

So if your plans are to provide private cloud computing to your users, or at least behave as one, you need to consider a service catalog very early on to help you establish standards, service levels, and provisioning processes.

This time, we ought to know one thing: No Catalog, No Cloud.

Tags: , , , , ,

Part 4: 10 Things Vmware Server Admins Should Know About Self-Service Catalogs and Lifecycle Management

Part 4! of the series “10 Things Vmware Server Admins Should Know About Self-Service Catalogs and Lifecycle Management” that I’ll be publishing over the next couple of weeks.

4. There’s a lot more to setting up environments than provisioning servers, the service catalog provides access to all the other required services.

It’s great that you can quickly get a server instance going. Awesome, really.  But that’s not an application hosting service. It’s important to understand the “whole product” requested. Is it a raw computing power w/ an OS? or is it an application stack? Or particular integration points, network, storage and security?

There’s a need to really think from a whole product perspective. What ancillary services need to go be coordinated to deliver an environment. Typically we’ll need to consider

  1. Network
  2. Storage
  3. Security
  4. Middleware
  5. Data
  6. Applications

Sometimes we can create complete application stacks such as: “Small Linux / JBoss / Oracle for standard development.” Other times, these items required hand-offs between teams.

In talking with VM admins, sometimes there’s a bit of the “not my problem” mentality — it’s those other jerks who are slow. But if the think about our job as delivering environments that can work in a data center or in the cloud and as we virtualized the network and storage, there’s more and more need for having a catalog of individual server request as well as complete environment

The service catalog contains all the other services that the customer needs in to deploy their application.

Tags: , , , , ,

Part 2: 10 Things Vmware Server Admins Should Know About Self-Service Catalogs and Lifecycle Management

This is part 2 of the series “10 Things Vmware Server Admins Should Know About Self-Service Catalogs and Lifecycle Management” that I’ll be publishing over the next couple of weeks.

Number 2. The service catalog is the place where your user can document (communicate) their request

Let me show you an example.

If you go to an e-commerce storefront and choose to look at Cisco UCS servers, they are broken down their servers into classes (Rack, Blade, etc), which then provides different models, which can then be customized within the parameters allowed for that model.  I’m not saying this makes sense for your environment, but the break down between classes, models, and then self-service configuration is a useful construct for thinking about your templates.

What are your standard classes of environment you provide?  Could it be production, development, QA?  What about models? Could those be on-line transaction processing, extranet, intranet HR, basic web server, basic database?

We would want to ask entirely different set of questions and configuration options for an extranet, high transaction database than for a personal development environment, wouldn’t we?

It’d also make our job much simpler and faster if we know what parameters were involved for that particular request.

The service catalog is key to enable your customers to:

  1. Discover what’s available me
  2. Guide me based on my high level needs,
  3. Help me compare models, then
  4. Assist me in customizing my configuration.

And of course all the tracking, workflow and life-cycle management that the service catalog enables.  This is what makes a service catalog different from a “web form front-end” to a help desk — automation is the big difference.

 

Tags: , , , , , , ,