Having outlined some of the structural issues facing CIO’s and providers of cloud service management processes in my previous blog on managing cloud infrastructure I am left with a real challenge.
As a critic of the old, over scaled, expensive and wasteful ESM/ITIL based service management systems, how can I now offer some constructive suggestions on a new way of looking at service management in the cloud?
Cloud service management strategies
1. Find a common language
- ITIL provides a meaningful and commonly understood language for all parties. We can use this language to identify the differences between a global public cloud (Azure, Amazon, Google, Rack Space et al) and a locally hosted private or public cloud service management model (DIY, All of Government, Datacom, HP, Gen-i, IBM etc).
- We can run through the ITIL framework step by step and identify where the global public cloud provider owns the service management environment; where it is owned by a local outsourcer; and where this responsibility is shouldered by the internal IT Department.
- This allows the service consumer to see and pay for the value being delivered while ensuring all other parties have a clear, profitable and sustainable model for delivering on their promises.
2. Be clear on the desired outcomes
- Rather than requiring that major public cloud providers redesign their internal processes (unlikely when dealing with players bigger than Australasia’s biggest companies) we should instead focus on the key elements we require. Namely:
- Agreed service outcomes. Put responsibility for various outcomes where they belong. For example: Responsibility for reporting and maintaining service levels may lie with your local service provider, while configuration, feeding and watering servers and storage would rest with the public or enterprise cloud provider. As you move further down the stack from SaaS to PaaS then IaaS, the role of either the local or internal provider becomes greater: that is, configurations, server OS updates and the like fall into the local role the closer you get to bare metal.
- Agreed method for support. It is important to note that public cloud infrastructure is already subject to 24/7/365 support, is highly available and is beyond your reach.
So how does the local IT team or its support provider engage with this engine to deal with the issues like change, incident, problem and performance management and service optimisation?
You (or your support provider) need to be able to monitor cloud services, respond to problems and engage effectively across the whole ITIL framework to optimise service and capacity as well as financial and system performance.
- Known Service Management/Service Desk cost model. You (or your support provider) need to be able to monitor and report on consumption of services. Without this ability neither party has the information required to determine the cost of any Service Management construct that based on a per unit charges.
By using the ITIL model as a frame and keeping our focus on outcomes we can assess what is ‘in’ and what is ‘out’ for each party operating the new model and to assign activities and costs accordingly.
In the diagram (above) the Service Delivery, Service Support, ICT Infrastructure Management and Security Management components can be provided by any or all of the three possible cloud providers – global public, local enterprise and internal ICT. No matter who is responsible for delivery, these components underpin the bits that really count: high value business, applications, processes and services.
There is now a clear distinction between what is valuable to both the consumer and the provider – and therefore requires active monitoring and interventions on a chargeable basis – and what is simply commoditised at the back end – the cloud providers’ service management responsibility.
4. Understand and enforce the distinction between how we use commodity Infrastructure and how we run and operate high value applications (Business Services).
As I mentioned above – the closer you get to the bare metal, the greater the role of the internal or local provider in such matters as configuration management and software updates.
As time goes on I expect most companies to move up the stack towards SaaS and for CIO’s to demand a greater level of simplicity from their own teams, software providers and developers to mitigate the cost and complexity of operating system updates, software upgrades and configuration changes.
This demand for flexibility and simplicity would mitigate against businesses accepting continued heavy customisation by developers and architects. These have traditionally become ‘sea anchors’, locking the customer in to a particular software product and making change almost impossible. Anyone using customised CRM, Supply Chain or similar database led services will know exactly what I am referring too.
We should see CIO’s insisting that their ICT departments move their focus away from designing, building and running bespoke infrastructure – currently subject to a race to financial zero between the big multinationals – and concentrating on the delivery of simplified applications and business services. This in turn will force infrastructure service management further into the hands of the cloud service provider with internal IT departments focussing on SLA’s, OLA’s, service performance, consumption and cost.
Identify the similarities and differences in the way we deliver and consume Service Management today and the new cloud environment. Having pin-pointed the differences, drill down into who owns and pays to deliver each component.
We can then establish a commercial framework that takes into account the need to achieve value for all parties. We can use the existing ITIL model as a way to frame this discussion and ensure we are using a common, meaningful language at all times.
In short, businesses should retain responsibility for Business Perspective, Business Processes and Application Management while a vertical line should be drawn between the service support (local provision) and the infrastructure management (cloud service provider). This in turn will drive a requirement for less customisation, greater simplicity and improved flexibility as we move up the stack towards Software as a Service.
If you have any thoughts/feedback, I’d welcome your comments.
The fourth and final part of this blog series of cloud service management will provide you with an insight into how to guarantee that you are getting the most value from your service management activities while staying tightly focused on the needs of your business and users.
Read the previous blogs in this series:
Find out more about how David and the team at Total Utilities can help your business with telecommunications and ICT requirements.