Cloud Migration; From Planning to Execution

Updated November 2026. This was first published on 21 October 2021. The idea is the same. Any number below is what the original piece said then. It is not a new count.

Key takeaways

The process of moving data centers, applications, and other IT resources to the cloud is referred to as cloud migration.
Cloud Computing is not a new concept and dates back to the 1950s when mainframes having superb compute capacity and storage ability were made available to multiple companies that could access the mainframe’s storage and computing ability.
The practice arose because the companies couldn’t afford to have mainframes with superfast computing and storage.

The process of moving data centers, applications, and other IT resources to the cloud is referred to as cloud migration. Cloud Computing is not a new concept and dates back to the 1950s when mainframes having superb compute capacity and storage ability were made available to multiple companies that could access the mainframe’s storage and computing ability.

The practice arose because the companies couldn’t afford to have mainframes with superfast computing and storage. In the 1970s, IBM took this practice forward by launching operating systems called VMs (Virtual Machines) which could run on a single physical node. From the 1970s till date, cloud migration has jumped multiple times, with companies getting multiple advantages.

The Major Advantages of Cloud Migration — Scalability- You can scale up and down resources #

Disaster recovery and backup- You can use cloud-based backup and recovery solutions that save your investment and time and use the cloud providers’ current backup and recovery what it can do that companies can seldom afford.

Task automation- Many orchestrations and automation of tasks like automatic updation in the cloud environment reduce the chances of errors and save time.

Though there are numerous others, let’s not delve more profound into it as the primary focus is to know the process of cloud migration. The overview provided is to justify the companies?

The entire process of cloud migration has been detailed out from scratch.

1. Planning For Migration #

There are various company-specific objectives for migrating to the cloud. The various use cases of migrating to the cloud can be — The urgent need to reduce IT costs as a supplement to cost-cutting for the entire company — Improve the security of the company IT assets — Induce agility in the process and speed up workflow — Improve productivity and availability of resources — Have better incident management — Accelerate the growth of the company.

Generally, the proposal for migrating to cloud comes from the CTO or the CIO of the company, after they have been apprised by different department and business units of challenges to be resolved.

2. Forming the CCOE (Cloud Centre of Excellence Team) #

The team should be built by drawing in the people involved in cloud migration from different departments and teams. It should always have top management inclusions or people from leadership like the CTO, CIO, CFO, and VP. would provide the team the authority for getting sponsorship. The other team members must be chosen from operations, infrastructure management, security management, and application management.

Note: *Most of the time, the CCOE leader tries to select versatile people in all areas of the cloud. That perfect team never exists, and the CCOE may go on hiring people with all the cloud skills. This bears a huge cost for the company without assurance that these people would truly deliver for the company-specific uses cases. The best way is to select specialists in their domain and having the necessary cloud skills to implement the minimum necessary quality environments on the cloud. And in the process of migration and planning, selective cloud experts can be hired to train and upskill as and when required. During the selection of CCOE, try to pull together diverse professionals excelling in different domains so that you are assured of the best frameworks, guidelines of operation, and reference architectures that may be required.

Major functions of CCOE #

( Reference image; It needs to be changed) #

The CCOE acts as a pivot for all activities pre, during, and post-migration. The team is responsible for — Onboarding of cloud providers — Framing the cloud computing policies — Selecting the technology and tool stack for reducing costs on cloud and governing in a better manner — Architect cloud solutions and ensure migration to the cloud successfully — Optimize the cost and risk aspects of cloud migration — Evaluate, assess, negotiate and select vendors who can deliver according to the company objectives — Ensure steady communication of all the cloud activities to the people involved.

Note: The CCOE shouldn’t try to enforce all the cloud activities collectively in one go. Enforcing these activities means changes in workflow, and the person in the company heading the workflow may resist such change. They may feel that they are being forced to make changes. So, it is essential to partner with the concerned department head or the person entrusted with the workflow and communicate how the changes would benefit the process. Make small, incremental changes at first and then slowly scale up the changes.

3. RFP for vendors #

One of the significant functions during cloud migration is selecting the cloud vendor. For this, a proper RFP with all the essential questions that can address the concerns of the the people involved and can meet the company objectives must be framed.

What is RFP for cloud vendors? #

The RFP for the cloud vendor is the request for proposal. This document enlists all the business requirements from the cloud vendor that would be required for cloud migration. For example, it may include questions like whether the vendor can make the cloud services compliant with necessary standards. The RFP is generally meant to check whether the cloud vendor can suffice all the company objectives of migration and operation on cloud.

Specific questions that you can ask in the RPF #

Detailed Requirements — What if any what it can do do you have to provide Professional Ser #

What what it can do do you have to provide Managed Services for our Infrastructure/Applications hosted in your Cloud? Is this offered direct or through a 3rd party and please provide details.

How would you help us mature our COE practice to keep up with your Cloud’s changes / new features?

What Licensing, if any, I can bring along to optimize my cost running in your Cloud?

What integration what it can do do you provide to integrate with my existing on-prem IAM (AD) solution? Will I have to rewrite my security posture related to IAM? Or will we be able to reuse the existing?

Are there any additional network costs that you have outside of what would be considered dedicated links (e.g., direct connect, express route, inter-region)?

What what it can do do you have to help reduce our bubble cost in migrating to the cloud?

What Cloud Training Curriculum do you have available that could be used to train our staff with your technologies and possibly integrate with our training platform Cornerstone?

What if any credits can be applied to our corporate cloud account to help facilitate our migration and real-time training using your technologies?

RFP Activities — Creating the RFP, taking the people involved on board while framing the q #

(Reference image- needs to be recreated) #

The SLA should contain — Statement of objectives (the purpose, goal and the objectives).

Information about the stakeholder’s involved — Scope and phases of cloud services to be covered by the vendor — The specifications of the cloud services to be provided by the vendor — Performance metrics- both resource and composite metrics (Response Time, Downtime/uptime, resolution time) and KPIs — Compliance and security considerations — Penalties if any of the party breaches the contract — Supporting processes, exclusions, limitations — Responsibilities etc — Reporting processes, indemnification clauses etc.

Few examples of SLA which may help you outline your cloud provider SLA are provided below.

The Core Activities for Migration #

Once the SLA is mutually agreed upon, the cloud provider undertakes the following activities — It undertakes detailed audit of the IT infrastructure, including applications and its dependencies — Creates the list of VMs, applications and other assets to be migrated to cloud — Use proprietary tools and other technology solutions for orchestrating and automating the process.

The entire cloud migration occurs in various stages — Deciding the objectives for migration (as has been discussed earlier).

The different cloud migration strategies that can be employed are.

Rehosting (lift and shift)-** It involves shifting your stack from on-premises to cloud.

Consider lift and shift hosting when you need to move to the cloud quickly, when the complexity of migration is less, when your critical apps are already operating in the cloud and you have defined migration pathways.

Replatforming**- Replatforming involves making few modifications for cloud adjustment of applications. The application core architecture remains the same.

Repurchasing**- It involves moving applications to new, cloud native product (ex: moving CRM to Salesforce).

Refactoring**- It involves modifying or rebuilding your applications from scratch to use the cloud what it can do.

Retiring**- While assessing your application portfolio for migrating to the cloud, you may identify some applications which are not viable for use on cloud.

Note: As a best practice, keep the dynamic data close to compute facility and the static data close to the user.

You can either migrate to private cloud, public cloud, hybrid cloud or multi-cloud, depending on the requirements for migration.

Listing the VMs, applications or workloads to be migrated to cloud.

A container is a software unit that packages up codes of the application and its dependencies. This package can run independently with all its dependencies. This OS level virtualization helps in distributed running of applications without deployment of the entire VM.

For containerization — You need to have the source codes of the application from repository — Install the tools, employ configurations for communications across applications — Ensure that the command lines of the codes could run with all its essentials independently.

Before live migration, the container is placed at the source node. After the migration, the container is placed at the destination node. To enable migration, the container is freezed at the source node by the platform locking up memories, systems, network connections and file systems. At the destination platform the container is unfreezed and the live state of container is restored.

Refactoring of applications to make them cloud native.

You can go in for many types of refactoring like.

(a) Complete Refactoring- Here 50% of the application codes are changed to have maximum cloud native use.

(b) Minimum Viable refactoring- It involves minimum changes in applications so that they are compatible for running on the cloud.

(c) Containerization refactoring- Here, the applications are made into containers with minimal modifications to improve portability that helps in cloud migration.

(d) Serverless refactoring- Here the applications are modified so that they can work excellent with the serverless systems on the cloud.

This is the test project where few workloads are established on the cloud, which acts as a proof of concept. The execution of pilot project helps identify gaps and issues that can arise in migrating all the workloads. The candidate is selected for the pilot project based on the assessment of the application or system’s usage of resources, connectivity, architecture’s compatibility with cloud, topology and others. You build the foundation cloud for pilot project, test the selected candidate. The pilot workloads are then migrated and serve as representative use cases. Migration scripts are formed for automation of the process. The model is then verified economically against the entire application portfolio of the company.

Here the landing zone for the workloads are established, right sizing of instances are done, configuration set-ups are enabled for migration.

(Reference image- Needs to be recreated) #

Testing pre, during and post migration with a shift left approach helps to mitigate challenges before time and reduce cost and time involved in resolution solutions.

All the necessary configurations set up for network, storage, compute, monitoring and others must be done to ensure that your systems run at peak performance levels on clouds without any downtime or disruptions.

Note: You shouldn’t wait at add monitoring services until migration completes. Put in place a solution that can define the pre-migration baseline, provide insights during migration and post migration. The solution should be able to collect all log files of the cloud provider and dish out the vulnerabilities and threats proactively. This enables end to end operational visibility and eliminates conflict on missing of the KPIs. The monitoring solution’s alerts and dashboards shall quickly identify and inform the issues before production.

Questions and answers #

What is the main point? #

The process of moving data centers, applications, and other IT resources to the cloud is referred to as cloud migration.

What else did the original piece say? #

Cloud Computing is not a new concept and dates back to the 1950s when mainframes having superb compute capacity and storage ability were made available to multiple companies that could access the mainframe’s storage and computing ability.