How COVID is Fuelling Cloud Migration
The COVID-19 has pushed many companies into a domino effect wherein they are accelerating the facilities for seamless remote work. Here, cloud computing emerges as an essential technology as it is capable of handling critical applications and at the same time scale the infrastructure based on the requirement.
Around the globe, the CIOs’ are the first responders in order to ensure that their teams are safe and comfortably placed to work. Next comes the business continuity and a major part of that is enabling their teams to work remotely.
It is interesting to observe how COVID 19 is pushing and accelerating many enterprises, both large and small, to migrate to cloud.
THE CLOUD MIGRATION JOURNEY DISCOVERY & PLANNING PHASE
It is very important for an organization to understand the interconnections & interdependencies of multiple layers of applications and servers so that it can decide which servers can be moved to the cloud. Here Automation would help standardise configurations for multiple server hiccups – compatibility issues with software especially operating systems; outdated OS that may need upgradation.
SELECTING THE RIGHT SERVICE MODEL
In these challenging times, it is essential to select the right service model and get a deep understanding of Cloud Adoption Frameworks of various cloud service providers. Depending on the type of applications which has to be moved to cloud & the kinds of workload – compute, storage, network – the organization can decide which type of service – IaaS, SaaS or PaaS is to be considered.
Not all workloads/applications need to move to the cloud. The 6 R’s approach helps in deciding if a particular application is/isn’t suitable for cloud environment:
Rehosting Or lift & shift:
Rehosting is the (currently) the most common migration method (40% of all migrations according to Forrester) due to its relative simplicity and speed. This method is the staple activity in a large legacy migration scenario where the organization is looking to quickly implement its migration and scale to meet a compelling event such as evacuating a datacentre or hosting provider.
Here, based on observations about the application, the organization makes an informed decision to change one or more components of the application, and in doing so, achieve additional optimizations. To be clear, changing a component does not change the business logic. It does, however, require changes to configuration, set up, tear down, and administrative practices.
This talks about replacing on-prem software with SaaS solutions. If an application is difficult to migrate to cloud for various reasons like upgrading, etc. then, going for SaaS may be a good strategy. E.g., CRM can be moved to Salesforce.com
Refactoring is a method to transform a non-cloud application into a cloud-native application. This method transforms everything about the application: it’s components, the application code, and the data itself.
This often consists of a targeted initiative to identify and retire traditional (legacy) applications from the portfolio. For large, enterprise customers, it is not uncommon for there to be multiple versions of the same application running in their environment, along with “orphaned” applications (those that have no clear business owner or sponsor and, in some cases, have no developer resource allocated due to attrition of the original developer or development teams).
Applications remain on-prem; depends on criticality of applications and prioritisation. This often happens when legacy OS and applications are not supported in the cloud and the business justification for migrating is insufficient.
- Perform a lift & shift of applications with minimal changes to ensure that the team is comfortable with the new environment.
- Have a Hybrid cloud strategy (private and public cloud) as a fallback option in case of contingencies. Start with least critical applications so that there are fewer disruptions.
- Use automation as much as you can as it saves time in terms of deployments. Anticipate compatibility issues esp. with operating systems
- Throughput on cloud may not be the same as on-prem. When the new environment is ready, initially, test with limited user load with limited users in real production environment; over time, shift 100% of workload to cloud.
- Once customers graduate from Phase 1, they go for a multi-cloud environment where they have 2 or more public cloud providers
- Adopt cloud native services like serverless architecture, spot instances, auto scaling, turning off servers when not in use – this will help achieve maximum cost savings, improve teams’ productivity and time to market
- Utilise cognitive services available on cloud which will give you immediate benefits – image recognition, automation, insights, knowledge search, etc.
To have a good Migration Strategy, it is equally important to have a team with right skills. It is essential to have management buy-in & ownership so that there is collective effort. Along with this, the team should follow DevOps Automation practices and use Infrastructure-as-a-Code while executing the migration.