IFR in the Clouds: True Heading & Cloud Computing

This is an interesting Oracle Whitepaper about Cloud Computing. Just published  in August 2009.

Title: "Achitectural Stategies for Cloud Computing". 

Good article in the German online magazine www.spiegel-online.de about Cloud Computing in Germany.

Sorry, no English version available. Main points:

- No clear defintion of the term CC make this trend hard to catch for most users/companies

- Most smaller companies worry about control of their data. This is the main difference to the US, as legal regulations especially for private data are significant higher in the European Community.

- About 90 % see CC not as an proven technology, they prefer to wait for 1 or 2 years and not to act as early adaptors.

This is somehow true, as I know from several of my customers that they prefer the role of a "fast-follower", as this is a good balance for them of being somehow state-of-the-art and not at the (expensive) b(leading) edge.

How about SAP, Oracle Databases and other complex architectures in the Cloud Space ?

SAP comes in two flavors in the Cloud Computing space. One is pretty simple to define and describe:

It is the typical SaaS Solution:

SAP Business ByDesign. Simple pricing model, but there is not too much of long year experience in the market, as this Software solution is pretty new....

We will here focus on the Standard SAP Business suite. These SAP Applications are a central part in larger companies. They build a design backbone for the infrastructure.

At this point, standardization is key to your envoronment. Use simple, but effective Operating Concepts. Use Solaris and/or Solaris x86.

You may find excellent information about SAP Infrastructure Solutions including detailed recommendations for setup of more complex environments like Oracle RAC at The SAP on Sun Wiki

 

Why are there no market prices for IaaS or other basic services in the Cloud Computing space ?

In my opinion Cloud Computing

• Is not a technology, but a market trend
• Dominated by non-IT Companies
• Revenue Streams out of  IaaS are there, but more or less irrelevant. Just too small.
• Most IaaS offerings are just a starting point

This is just a foil out of my latest and greatest CC presentation. It is a simple sample for a logistic company.

Better:  For a logistic company with own IT Services.

Here is my recommendation -)

1. Start offering cheap IaaS services for dedicated customers.
2. Add Web content including design services
3. If your customer is lucky with your services, start offering additional services like:
4. A Catalog Management System for the products of your end-customer
5. Offer a Price list Management System. You will need this, as you will offer flexible catalog systems.
   This requires dedicated catalogs or shops for different customer groups: Different prices for end-customers and for
   Retailers.
6. Well, if you know the customers, you can offer services that will cover peaks. Offer campaign management, for seasonal or other peaks.
7. If you already manage the Campaign, enhance your offering: Ask for the complete campaign logistics. If your customer wants to
   sell 400.000 footballs in 2 months......it is a good idea, to outsource the logistics for this peak.
8. If you do your job good, you may offer to take over the complete logistics for your end-customer
9. This may include not only stock/warehouse, but call center and / or CRM Systems

This is just a logical order......There is no starting point Cloud Computing.

Remember: Cloud Computing brings the idea of book-shop-logistics to the IT Department.

• There is no such thing as a isolated IT Cloud Solution. It is all integrated in one big business scenario
• There is no market price for IaaS. It will be more and more combined, merged, mashed with additional offerings
• These additional offerings will be the true heading here. This is where the margins are....
• Remember: No or low margins with IaaS. This is  the land of  sharks and desperates......
• This is your big chance: This ties your customer to your company. There is no loyalty with IaaS only.
• If you can add/combine your business with IaaS or SaaS solutions: Cloud Computing is your technology.
  

Now, here is a sad story.

I walked into that at one of my last CC presentations.  The customer was already faced with some negative effects:

• Parts of their business migrated already from reals life to the web space.
• Parts of IT Services are alredy lost to Cloud Computing suppliers (SaaS) Remember: The pareto about the revenue streams !
  

At every customer workshop we do the same brainstorming when we talk about the pareto:

How can IT services be expanded, how can they be part of a more complex solution. How can we rise the contribution of the service ?

At this point we failed. They recognized, that they could not extend or enhance their offerings to their end customers.

They, on the other hand had almost no chance to extend their offerings in the pre-production area.

We all noticed just in that situation: The situation is more serious than expected.

So, now you know why I thing that Cloud Computing is not about IT Technology. It is more like SOA, but spreading from one company to the next.

From IT Technology to business process and more into business. It is 2009, Cloud Computing is just starting to change. And the extreme cost pressure from the economic crisis will drive this faster and faster.

Why not use Sun Solaris when you migrate legacy Applications into the Cloud ?

Instead of  rising the complexity with

• Maintaining one Operating System
• Maintaining an virtualization layer like Xen, Vmware etc. with an included OS

why not reduce the complexity and

• Run Solaris or OpenSolaris
• Use ZFS as a cool Filesytem/VolumeManager
• Use Solaris Container to get the flexibiliy you need....

Find more information about why Solaris

Good advice, how to to simple and secure patches sorry until End-of-Mai in German only

View the comments to get all the information you may need to run a complex, virtulized environment without any additional tool.

REDUCE Complexity is the first rule :-)

On the same Wiki you will find some more generic information on "Why you should run an SAP Environment on Solaris instead of Linux"

Meine Empfehlung für die Nutzung von Cloud Computing Services:

• Beachten sie die rechtlichen Anforderungen an den IT Betrieb
  
• Holen Sie fachlichen Rat bei einem Anwalt für IT Recht und / oder einem IT Sachverständigen
• Dokumentieren Sie die vollständige Service Nutzung. Wichtige Hinweise hierzu bekommen Sie durch  ITIL V3

Und keine Sorge: Bei Beachtung der rechtlichen Rahmenbedingungen können Cloud Services durchaus ohne größere persönliche Einbußen genutzt werden :-)

[Read More]

Zum Schluss die Empfehlungen:

• Behandeln Sie Cloud Computing wie jeden anderen Supplier
• Beachten Sie, dass Cloud Computing einen h
  
• Planen Sie Alternativen/Fallbacks, falls der Cloud Service nicht weiter angeboten wird (EVAPORATION)
• Zu jedem Supplier-Vertrag gehört eine klare EXIT-Strategie
• Ignorieren Sie die Aussage unsinninge Aussagen wie "ITIL und Cloud Computing funktionieren nicht miteinander"
  

[Read More]

If there is a part one, there is a part two.

in the first Long Tail blog entry I  described the influence of Cloud Computing to the revenue streams of a ISP or an Internal-IT / data center.

Today we will discuss the usage of Cloud data store, like Amazon S3 services, using the same Pareto model.

The typical pricing model for a data store is a combination of

3 different pricing factors:

1. Price per GB uses. This is somehow in a range from 0.180 $ (Dollar ?? for Europe ??)
2. Price per GB transferred out. Range from $0.170 per GB – first 10 TB / month data transfer.
3. Price per request. Range from $0.012 per 1,000 PUT, COPY, POST, or LIST requests

Here is my second Long Tail recommendation:

x-axis : Describes the amount of computing = CPU power a given service/application needs per storage request. To the left we find applications with almost zero CPU requirements like file services. To the right applications get more and more CPU bound, like ERP, Mail etc. 

y-axis : This is the combined costing out of Storage cost, data transaction cost and service request cost (PUT,GET etc.).

The higher, the more total access cost will occur when the application is in use. It is the cost function out of the  3 different pricing factors.

The Head = The Red Head

These are data that are frequently used and that require high up/download bandwidth.

Typical scenarios are:

• videos
  
• high quality picture / photo storage
  

The Long Tail = The Long Green Tail

Here you find data that is probably not used frequently or that has limited storage requirements. These applications are in general a good fit for the Cloud Space.

My recommendation

• Keep in mind that there are 2-3 price components
• Keep data with very high access rates in your own data center
  
• Focus on the Long Tail. The Cloud is a perfect place for this kind of data
• Think about legal or Compliance rules. This is important for Europe, probably less in the US.
• Keep an eye on the workload distribution
  

Other influence factors

Workload distribution is the next big decision factor. If you have a quite good balanced access to the data during a month, the above recommendations will

work fine. The infrastructure cost are somehow balanced and you only need to keep out of the red = Head area.

Things can get different, if you have  workload hot spots. Or if Hot spots are your main business. On CloudCamp I had the chance to talk to companies that have only

a few hot spots per year. Some web-pages are only created for special events like the Olympics or other sports events.

This is a clear indication for high flexibility requirements: Use Cloud Computing. No question.

Thanks to Jan Brosowski, who helped me get the x/y coordinates right, as I was lost in Pareto space.........

You are an ISP or an internal IT Provider in a large corporation ?

Here is a good reason, why you should care about Cloud Computing. The Cloud will attac your datacenter.  Your datacenter will be flooded. You are doomed.

Well, we are a little more optimistic here. We take a more detailed look into your datacenter revenue/earning stream:

The Head

To the left, the red area,  you see all your services that make only little revenue per incident, but usually appear in very large numbers.

A good example for this kind of applications are:

 - File and disc Services aka user storage

- Mailboxes

- Backup space for endusers

If you are CIO, you will make a good amount of revenue or cost coverage with these kind of offerings. Every user needs them. If you run a datacenter supporting 200.000 employees, these basic infrastructure service will be a big deal. Easy to manage, long term revenue stream. Small revenue/cost coverage per unit, but high ammout of users.

As these applications require lots of servers, storage etc. they give you the purchase power. Every hardware vendor is very pleased about large numbers. You need them on the hand to optimise your cost structure and on the other hand they still give you a good revenue stream.

The right part aka The Long Tail

These are the services that are used and requested  only a few times in your datacenter. Samples here is a Oracle RAC installation, or a warehouse application running on 3 machines that require a full audit or a FDA validation. The service is only rarely, they are complex and expensive to set up and maintain. But they still are important for you overall balance sheet.

Low purchase pricing (caused by The Head) combined with high margin Services from the Long Tail.

How is the Computing Cloud moving into you datacenter ?

Remember my blog entry "climbing climbing" ? In this order external Cloud suppliers will enter your organization. Endusers or whole subsidaries will get offerings for easy web services, sometimes cheaper than anything you have ever seen. I know examples, where a large SW vendor offered a complete hosted mail platform in a Cloud cheaper than the licence cost for the same amount of users. And this is just the beginning. Remember: There are no market prices for Cloud Computing services. All prices are, let's say marketing / political / strategic etc..

Whatever you do as a ISP or Internal-IT department:

The Cloud Computing offerings come with extreme low price and margin. Because all suppliers know how to act in a low margin market.

This is the point where things get ugly for you.  The "red" area aka "The Cloud Target" will  erode. The easy but high volume services will leave your datacenter.

This will cause two main effects:

• As your purchase volume shrinks, your purchase power is reduced and your average cost will rise
• Your overall revenue will shrink, with negative effects on cost degression

 But as all the more complex services still remain in your datacenter the most cost effective services will remain. Double trouble.

Do not worry about the Long Tail. As these services are complex, have high requirements for validations, testing or compliance:

These services will never be hosted in a Cloud. At least not in Europe. My bet.

OK, got it. What can I do ?

• Take Cloud Computing serious
• Make your own Cloud offerings
• Run a Private Cloud for easy services
  
• If your customers, internal an external request services from a Public Cloud: Provide the contract. Remember: You must not allow somebody else to step into your contractual relationship with your endcustomer. Using Cloud Services from one of your subsidaries may be acceptable for you if YOU own the contract.

The internal IT department and the Cloud

At one of my last customer visits Cloud Computing was not very popular...

The usual refusal:

- That will never work !

- This is only a 3 quarter trend..

- Aah, by the way, we are doing Cloud Computing since 5 years.

The last ones are my favorite ones....You talk probably to the slowest moving and lease inovative  ISP in the german market. Maybe it is caused by their company history or just by inertia:

All customer negotiations need months, their end-customers call them slow and unflexible.

 And than you are forced to listen to one person of the marketing department saying " Cloud Computing is nothing new. We do this since several years"

 Or, even better:  Products that are sold on the market since more than 10 years are now called Cloud Computing Security Solutions, claiming that they are your only chance to have security in the Cloud Computing space.

This is in my opinion a real danger for the Cloud Computing idea. Same happens in the Green-IT space. After a few months every server is green.

This kind of non-sence treating you as a customer as a fool is called 

Cloud Washing =  Repainting  standard offerings as new, trendy and Cloudy

So, you just take your standard product offering, print "Cloud Computing inside" on the bag an off you go ! If you work in the product marketing department of one of these company, here is the perfect tool for you !

What are my recommenations for ISPs or internal IT Departments ?

1. Be open. Open skies.
   Take the opportunity to learn about Cloud Computing. Take it serious.
   
2. Manage risk, manage compliance requirements
   There are risks around. Vendors might disapear (Cloud evaporation). In Europe you have to deal with high Compliance requirements, even rising.
   
3. Manage vendor-lock
   Don't use a Cloud Computing offering without a clear EXIT strategy !!!
   
4. Offer cloud services, no Cloud Washing
   Offer your customers easy and new Cloud Computing offerings. You don't have to start with 100 Servers or 10 SaaA. No Cloud Washing, never. Your customer will notice and will hate you.
   
5. Consulting services
   Offer consulting services around your new business. ALL of your customers have high demand on technology neutral consulting services, thats for shure. Offer your services fast, or your customers will listen to somebody else.
   
6. Architectural services
   Integrate Cloud Computing offerings into your standard service catalog.
   
7. Cloud services, but with acceptable SLAs
   Offer services only with acceptable and good SLAs. If you want to know what is not accpable: Just have a look at the SLAs of most of the Cloud Computing Infrastruture offerings in the web/cloud.
8. Strict control # of operating systems
   Reduce the number of operating systems and operating system variants. Keep in mind that a virtulized operating system is still an operating system. Remember that  a virtualization SW does not solve your Release-/Patchmanagement requirements. Learn about the best operating system for cloud computing.
   
9. Cloud services must be part of a solution
   Don't even think about IaaS offerings. This is a crazy market with no market prices but political pricing. You as a small company will have no chance in this market, beleave me. Offer allways more complex services like PaaS or SaaS.
   
10. Offer building private clouds
    Just start your own little Private Cloud ! Read how to start your own little Cloud in one of the next blog entries :-) stay tuna.
    

When your perception is cheeting you..........

Cloud Computing is complex, the internet is involved (thats the place where all the hackers life), your data are far away (you can probably not visit them in the data center and take the disks home).

On CloudCamp one of the artists called this way:

"You have a faster, better, more flexible and more state of the art solution. And nobody does understand it"

As with every new technology, every new approach there is a certain subjective perception of danger and trouble. Especially, when it comes to control over data.

We have to discuss security in general to get a better understanding of the underlying problem.

Data corruption, loss of data and issues with data integrity

Can have both internal and external sources. During Identity Management and Security Projects we found internal sources in a lot of cases as the root cause of fraud. Some experts call internal fraud as the reason for data loss / data integrity problems with not less than 50% chance. So, the problem is maybe behind your firewall.

One of the recommendations to solve this is a strict security management with data encryption for everything in the network and/or everything on disk.

The same rules are valid for any kind of date stored in the cloud: encryption and strict access control.

Cloud Evaporation

A serious problem, as  we see a consolidation trend from smaller to larger Clouds. You can get around this problem, when you follow standard rules of engagements. Have a look in frameworks like ITIL V3: There are such things like Vendor Management. There are strict recommendations, that you should have a clear, proven exit strategy before you start using a service. It is so easy and sweet like honey to just start to use a service without planning a complete lifecycle for the service. This is unprofessional, the negative effects will kill some of the early adaptors.

A joke (from CloudCamp)

You order a pizza via the internet. The complete transaction will be done via https. But you won't type your credit card number, as this is not secure. You will call them an provide the number by phone.

But we all know that phones can be wiretapped a lot easier than an https internet connection. This shows the perception problem in a very clear case...

How should I use Cloud Computing Services ?

Where can we start ? Here is my personal hitlist.

OK, some people do complex services today already in the Cloud. In their eyes this ranking is wrong. Good expample is the massive use of salesforce.com.  For me it is just a good example of taking risk. The main risk is the exit cost or the vendor login. If you ignore this issue, solutions like that are fine already. But only if you completely igore Vendor Lock. Read here the ultimate interview with Richard Stallman about the Vendor-Lock problem. For me, the exit costs or even the chance to exit a Cloud is one of the key issues. What is the difference between a traditionall 5 year ISP contract compared to a Cloud that can't be left ? There is no difference, it is the same unflexible situation, maybe with some more risk. We go into more details in one of the next

1. Simple centalized storage for unstructured data like fotos, documents etc.
   
2. Backups for simple, unstructured Data
3. Web content, Blogs etc.
4. Email Services
   
5. Easy test and development scenarios
6. Short term campains for large Enterprises
7. Hosting for complete Web-sites
8. Dynamic WebSites, including Databases
9. Dynamic WebSites, including Databases and complex processing, ERP like applications

[Read More]

Let's start with a definition of Cloud Computing. When I do a presentation on CC, I start with an description of
Cloud Services, like Iaas, SaaS, PaaS. And, to get the ultimate definition, we ask Wikipedia:

Definition #1 Wikipedia (Version of March 22nd 2009)

Cloud computing is Internet ("cloud") based development and use of computer technology ("computing")....
It is a style of computing in which dynamically scalable and often virtualised resources are provided as a service over the Internet.....
Users need not have knowledge of, expertise in, or control over the technology infrastructure ......

OK, if you don't have control over your technology, this is Cloud Computing, I like that part :-)

But let's be serious, here are the main points:

- dynamic (or elastic)
- scalable (or elastic)
- virtualised (I don't agree, but this is an own blog entry)
- over the internet

Definition #2

Cloud Computing is technology term pushed by new-IT infrastructure providers to extend their business and market share in the traditional
IT space.  


The IT market is dominated today by hardware vendors like Dell, HP, IBM, Sun Microsystems, internal and external infrastructure providers.
Non-IT companies like bookshops, search engines or software vendors start enhancing their business by using their IT infrastructure and
know-how and offer IaaS, PaaS or SaaS services.

They do this for three reasons:

- They expand their traditional business (I will cover this in a next blog entry)
- Hardware vendors fail to offer flexible billing models ("capacity on demand", finance reasons)
- ISPs don't like the idea of non-bindig  business relations (old business models ?)

OK, the first definition sounds a lot better, as we are all technology and feature driven. Probably the second definition is the right one
if you watch Cloud Computing from a pure business/marketing view..

........ . The first sentence in a blog is very hard to write, so I just start with the second.

Too many blogs about Cloud Computing ?

Probably YES.

So why do you start a new one ?

Well, after visiting CloudCamp I noticed a very special demand in the Cloud Space: Most CloudBlogs are around Witchcraft and Technologie, about Interfaces, APIs Scripting Languages and, when it comes to Sun: About Servers, Storage and Q-Layer.

I miss the data center view on Cloud Computing. Business aspects, Legal aspects, how to make money and how to save costs. At CloudCamp we discussed questions like “Is there a difference between System Metering and System Monitoring ?”. A question we answered in the old data center world about 10 years ago.....Let's combine this experience with the enormous flexibility of Cloud Computing.

CloudFront and IFR

The Cloud front is moving from the internet closer and closer to the data center. And if more applications will be moved from “behind the firewall” into the Cloud...we have to enhance Cloud technologies with the ideas, experience, rules, regulations (sorry) of a classical data center. If you don't follow the exact rules and quality, your next flight into the Cloud will end with a ugly session with your auditors (And The Evil Eye Of The Hurricane's Coming In Now For The Kill). IFR rules will keep us alive :-)

This is my first entry. Please let me know if I am right or wrong here :-)   

Just hold your heading true.  Ralf