How to efficiently manage a hybrid cloud?

Today, any business that you look at is modernizing something somewhere. And, the demand is do it with minimum operating cost but ensuring a rich customer experience. So, what to do – opt for a cloud-based solution as there is no CAPEX involved, the technology has matured and is easily available.

I have been working with cloud few years now and, what I have witnessed is that most cloud services are managed in silos on a variety of platforms. I feel there is an urgent need for both the business and IT management that manages computing resources, whether in public or private clouds or in the data center, must be managed in a unified, consistent, secure and predictable way to secure business availability. For this we need a well architected, holistic approach to hybrid cloud management. While there are many tasks that must be accomplished to achieve hybrid cloud management, there needs to be automation and best practices at the core.  A good hybrid cloud management can satisfy business demands, operational excellence, cost optimization and IT governance.

Most people know what hybrid cloud management is on principle but how should you go about implementing it? First, understand the essentials. It is not just about randomly picking tools and technologies you want to use. IT operations managers make the mistake of focusing on tools that may make hybrid cloud management easier, without keeping in mind the business requirements – consequently, they get both the approach and the tool selection wrong. It is essential to understand the security, data, governance, and end-user dynamics since all of these will influence how you approach your hybrid cloud management. So, have business-based IT principles and a good architecture in place before getting into tools, automation and other discussions. I have outline 6 aspects of Hybrid Cloud Management for you…

#1 Understanding the nature of the Business

You need to understand what the applications do: how they interact with the end users, manage data, how they handle networking, security patterns, performance, etc. For that you need to know the following:

1. End users access locations, types of workloads (standard/non-standard) and business requirements
2. IT business and functional structures (example, manufacturing and Logistics, product team, R&D, etc…)
3. Decision criteria for workload placement
4. Good prediction on capacity and demand forecast specific to on-premise platforms
5. Configuration and compliance management along with reporting
6. Build or leverage existing standard operating environment
7. Strong integration between multi-stack systems

#2 IT Security Reporting and Compliance

We need robust security models to avoid business loss and one of the best recommendations is treat every service is unknown and introduce security gates in all the places and allow only what is needed. – “I think this is applicable everywhere and even in our daily life” there are two key aspects among other which I would consider always.
IT Security Directive rules: Organization should have the IT security guidelines and policies based on their business industry. This guideline and policies should be incorporated part of the standard workloads as SOE template enforced by the automation framework based on the workloads type and this workload must be assessed and report must be generated and reviewed by ops team periodically to ensure compliance on the internal/external auditing tasks. Addition to above, based on the policies, corrective actions should be triggered automatically and it’s good to have predictive analysis based on the environment recommendations.

#3 Unified, consistent self-service capabilities

Most vendors today produce their own native API for their products and indeed it works well when you look at from product perspective but not really from the hybrid cloud management point of view. There are few products/tools set, which abstracts the information from Software/Hardware and cloud vendors, but they are not fully matured yet to follow just install and use concepts.

Consider the following key points when we choose the toolset for this capability

• No vendor lock-in
• simplified catalogs view
• IT manager view for each one of the IT functional streams
• Leverage quota-based systems
• Tools integrations
• Leverage cloud/other private cloud vendors APIs & abstraction
• Multi-tenant models
• Charge and show back
• Simplified approval model for control
• Good authentication mechanism and review
• Security event reporting & automated actions

#4 Service models and Business commitment

In the world of infrastructure, workloads order and service availability has been known as service models and the service models can follow their terms (for instance, platinum, gold, silver and bronze) and based on this term along with additional policies will define workload placement strategy.

The key point is that service models contain the workload specification, in terms of SLA, hosting area, backup, recovery and associated components. This specification and commitment are very important for the application team. There wouldn’t be necessary that commercial discussion involved internally in terms of penalty but with vendors/suppliers it should be understand & agreed.

#5 Tools Nature, Integrations & Automation

One tool will not satisfy all the requirements and provide service richness. But there are key aspects needs to be considered when we chose tools, easy management, skills should be available in the market, broader coverage, should support process, platform support management.

1. Automation is the core of hybrid cloud management to achieve the cross functionality to produce usable services or workloads.
2. The more tools you use you will get more functionality and things will flow in automated way but this leads another challenge to face that integration between the tools which becomes complex, somehow all this integration must be documented & automated all the way.

#6 Process and culture change

People and process hold the key to success. If people and process don’t change to support “evolution”, you will not taste success. Technology is just an enabler, but people are the real key to make it happen. I don’t have to tell you what makes a difference as this is big concept and there are many working methodologies available to adapt which needs to be assessed and adapt it.

Published by

Ajithkumar Jesudurai
Cloud Solution architect
LinkedIn I

Ajithkumar Jesudurai has 9+ years of experience in infrastructure spend and is skilled in IT infrastructure management operations, service improvement plans, architecture and implementations of environment based on the business requirements. He loves to work on the multi-cloud setup and simplify the IT requirements in terms of catalogs (automation) and enable IT4IT services for everyone in a controlled way and improve IT business efficiency. He strongly believes, good architecture along with a good team always leads to fruitful result, no matter the level of complexity.