Virtualization Technology News and Information
Article
RSS
CopperEgg: The State of the New Cloud Supply Chain and SLA Dependencies

 

What do Virtualization and Cloud executives think about 2012? Find out in this VMblog.com series exclusive.

The State of the New Cloud Supply Chain and SLA Dependencies

Contributed Article by Scott Johnson, co-founder and CEO, CopperEgg

It is clear that a growing percentage of business processes and applications that have been traditionally hosted within a company are being provided as online services. Software as a Service is no longer a future possibility; it is commonplace.  The same holds true for Business Processes as a Service. Often these services are composed of a number of other components services. Which stands to reason ... it would be foolish for every service vendor to implement yet another billing system, when they can integrate with a billing service that is mature and inexpensive. My prediction for 2012 and beyond is that service vendors will need to address an as-yet unsolved problem: how to define, deliver and manage service quality guarantees for their own customers, when the service they are delivering is dependent on some number of other services from a variety of vendors, each provided with its own SLA.

This is the new world of managing the business service delivery supply chain. When I first realized the significance of this issue, I began calling it ‘the cloud supply chain'. This is not completely correct. The service delivery supply chain becomes important whenever delivering a complex service; that is, a service composed of or dependent upon services provided by other vendors, whether or not cloud computing is involved. The truth is, cloud computing will transform this notion from an interesting curiosity to a billion dollar market in the next few years. Hereafter, I will use the phrase ‘cloud supply chain;' its easier to remember.

How did we get here? I think the evolution occurred roughly as follows:

Software development practices.  As software development evolved, the notion of code reuse became an important paradigm. Write a function well once, and you would end up using it for years. Function libraries were born.  Functions evolved to objects, libraries evolved into beans and gems. With each step forward, mature components were used to build the next generation of components, each generation being broader, more abstract and easier to work with.

The open source revolution. The evolution described above was really made possible by the open source communities around the world. As open source communities grew, the process of building new components from existing components accelerated. Today, newer programming languages used in conjunction with widely available, mature open source and commercial components enable a single developer to construct an application of immense complexity in a few weeks. Well, less in an agile shop.

Today, no application is written ‘from scratch.' Applications are often composed of a number of other applications integrated together, along with many newly created components, which of course are built upon many, many other components that have been around for years.

Application integration and ‘middleware.' As applications proliferated in the data center, means were developed to share and exchange information between them. Methods and tools for monitoring the performance and availability of these applications were developed. 

Services delivered by IT.  In this discussion, I have focused on applications running within the same organization, which together deliver ‘services' to consumers internal to that organization. These services of course are also made available as to external consumers, as paid services. As described, these services are delivered by the IT organization, which is responsible for the quality of the service delivery. Every component of every application of each of these services is controlled by one group of people, who are responsible for the quality of the service delivery.

Stated another way, the IT department is delivering services which are built upon underlying applications. The building blocks of a service are applications.

Today we are entering into the next evolutionary stage ... where services are yesterday's applications. Meaning that services have become the building blocks upon which new services are built! In the case of a vendor that provides a service that depends on, or is built from applications and other services, if all of the constituent components are all sourced and controlled by that vendor, the management of service quality is no different than is has been for years.

Things become exponentially more complex in the world of cloud computing. Consider now a vendor that delivers an internet service, let's call it the ‘primary service', and provides a service quality agreement to the consumer of her service. Suppose this service is built upon three other services, each provided by different service vendors, integrated with a new application. This primary service is what I was referring to earlier as a ‘complex' service.  The quality of service delivery of a complex service is in-part dependent upon the service quality provided by the component services. In the case of the primary service vendor, said vendor has no control over the service quality of the component services. The primary service vendor must rely upon service quality guarantees from the component service vendors.

We have now returned to my original question: how does the complex service vendor define, deliver and manage service quality guarantees for their customers?  There is no simple answer to this question.

The following is a list of what I believe to be best first steps to take when you are faced with this question:

  • Decide upon the service quality level that your customers require.
  • Devise a means of monitoring your service for compliance with the SLA that you provide, emulating the service as seen by your customer as closely as possible;
  • Research and understand the SLA's from each of the component service vendors, and devise a means of monitoring each of these component services for SLA compliance.
  • Develop a troubleshooting methodology for when your service degrades; you need to be capable of quickly determining whether you need to address an issue in your own service delivery, or deal with degradation in the quality of a component service.

From a much higher perspective, I am simply pointing out that managing the cloud supply chain will become an important element of any business that seeks to market a cloud-based service. Management of the cloud supply chain is similar to traditional supply chain management. Many of the thought processes are the same. You need to screen and select component vendors, and evaluate their products. You may consider second-sourcing critical components, whether in the traditional sense of a second vendor, or in the sense of providing a second delivery source (cloud). The supply chain management process has been evolving for a very long time, and there is a great deal of wisdom that can be gleaned from that field of knowledge that is directly applicable to cloud supply chain management.

The single biggest difference between traditional supply chain management and cloud supply chain management is time. An internet service is generally speaking real-time. There is a customer of your service waiting for your service to deliver something. A disruption in a dependent component service must be detected quickly, and the necessary corrective actions kicked-off within seconds. If you wait for the customer to notify you that your service is slow, or even down, you will be handing the advantage to your competitor.

At CopperEgg, our service is dependent on a number of other services. We use external secure billing systems, we purchase services from several cloud vendors, and portions of our real-time alerting system are dependent on SMS and paging services delivered by other vendors.  In order to insure that our real-time monitoring application performs to our expectations, we need to monitor it. We took the time to understand the SLA's of our suppliers, and we monitor their services for SLA compliance.  Our own cloud supply chain is vitally dependant on SLAs from numerous vendors; in-turn, several of our customers deliver services today to their customers that are dependent on the quality of our service delivery. Our SLA with our customers is in-part driven by the SLAs provided for the component services upon which we depend.

In 2012, I believe we will see a higher level of scrutiny demanded for each participant in the cloud supply chain.  New service vendors will need to publish their availability or lack thereof to their customers / consumers.  The interdependency of these SLAs is a concept that will need to be grasped and managed.  The idea of alternate suppliers will come into play as well if a particular vendor cannot meet demand, ala SLA.  My crystal ball displays that in the 2012 cloud, the problem of managing SLAs from multiple vendors will prompt a great deal of interest in cloud supply chain management. Tools and processes will evolve as service vendors work together to address this issue. Monitoring SLA compliance will become commonplace, at least in the backend, as a means for service vendors to interoperate and to make their services even more reliable.  I predict that the air of excitement and cooperation about the cloud will continue in 2012, and we will see the continued success of cloud vendors who are transparent, provide groundbreaking new services, and support their customers as if their businesses were at stake! Yes, my head is in the clouds!  Best wishes for 2012 to all, from all of us at CopperEgg.

###

About the Author

Scott Johnson is the co-founder and CEO of CopperEgg, which is the third company he has co-founded in the technology space (his first startup being in 1984 when he co-founded Thomas-Conrad Corporation with Walter Thirion, which was later purchased by Compaq). A passionate entrepreneur, he is committed to building a company whose cloud monitoring products are valuable, affordable, accessible and useful.

Published Wednesday, December 14, 2011 6:20 AM by David Marshall
Comments
VMblog.com - Virtualization Technology News and Information for Everyone - (Author's Link) - January 4, 2012 7:06 AM

I'd like to personally welcome each and every one of you to the start of 2012! As we begin what will certainly prove to be a fantastic new year, I wanted to make sure to thank all of the loyal member's and readers of VMblog.com. Once again, with the help

To post a comment, you must be a registered user. Registration is free and easy! Sign up now!
top25
Calendar
<December 2011>
SuMoTuWeThFrSa
27282930123
45678910
11121314151617
18192021222324
25262728293031
1234567