Cloud adoption has transformed enterprise technology.
But for most organisations, the destination is not "everything in one cloud."
Enterprise environments commonly contain a combination of:
- On-premise infrastructure
- Private cloud
- Public cloud
- SaaS applications
- Colocation
- Edge infrastructure
- Multiple cloud providers
The challenge is no longer simply adopting cloud.
The challenge is managing a distributed technology environment without creating uncontrolled complexity.
Start With Workloads, Not Cloud Platforms
Cloud strategy should begin by understanding applications.
For every workload, consider:
- Performance requirements
- Availability
- Data sensitivity
- Compliance
- Integration
- Latency
- Scalability
- Recovery
- Cost
Different workloads may require different infrastructure models.
The objective should be appropriate placement rather than maximum cloud adoption.
Why Hybrid Cloud?
Hybrid cloud combines on-premise or private infrastructure with public-cloud services.
This can allow organisations to retain selected workloads locally while using cloud for other requirements.
Potential use cases include:
- Application modernisation
- Backup
- Disaster recovery
- Development
- Analytics
- Burst capacity
- Global services
But hybrid infrastructure requires connectivity and security across both environments.
Why Multi-Cloud?
Organisations may use multiple clouds because different platforms support different applications or capabilities.
Multi-cloud can also emerge through:
- Acquisitions
- Department-level adoption
- SaaS requirements
- Customer requirements
- Geographic expansion
Without governance, however, multi-cloud can become difficult to manage.
Connectivity Comes First
Applications cannot operate effectively across environments without reliable connectivity.
Architecture should consider:
- Bandwidth
- Latency
- Redundancy
- VPN
- SD-WAN
- Dedicated cloud connectivity
- Routing
- Monitoring
Network design is therefore a central part of cloud architecture.
Create Consistent Security
Every cloud platform introduces security controls and identities.
Without a common strategy, security policies can become inconsistent.
Enterprises should create common principles around:
- Identity
- Privileged access
- Network security
- Data protection
- Encryption
- Logging
- Vulnerability management
- Compliance
Cloud security should integrate with the wider enterprise cybersecurity architecture.
Protect Data Across Environments
Moving applications into cloud does not eliminate data-protection responsibilities.
Organisations should understand:
- What data exists
- Where it is stored
- How it is backed up
- How long it is retained
- How quickly it can recover
- Where recovery copies reside
Backup and disaster recovery should cover the entire hybrid environment.
Control Cloud Costs
Cloud is highly flexible—which can also make spending difficult to predict.
Governance should examine:
- Unused resources
- Oversized instances
- Storage growth
- Data-transfer costs
- Reserved capacity
- Licensing
- Duplicate services
Technical architecture and financial governance should work together.
Standardise Operations
Infrastructure teams need visibility across multiple platforms.
Without centralised monitoring, incidents can require teams to investigate numerous consoles and providers.
Organisations should consider unified processes for:
- Monitoring
- Incident response
- Change management
- Backup
- Security
- Capacity
- Reporting
The objective is not necessarily one management tool.
The objective is one operating model.
Avoid Unplanned Cloud Sprawl
Cloud adoption frequently begins project by project.
Over time, this can create multiple accounts, subscriptions, tools and architectures.
A cloud governance framework should establish:
- Ownership
- Architecture standards
- Security standards
- Naming conventions
- Identity policy
- Cost controls
- Backup requirements
- Approval processes
Governance should enable cloud usage rather than unnecessarily restrict it.
Build an Exit Strategy
Every workload should have a lifecycle plan.
Ask:
- Can this workload move elsewhere?
- How portable is the data?
- What dependencies exist?
- What would migration cost?
- What happens if the platform changes?
Architecture that ignores portability can create future constraints.
Hybrid Cloud Is an Operating Model
The most successful hybrid strategies view infrastructure as a connected ecosystem.
Data centre, cloud, network, security and recovery decisions must support one another.
The goal is not to eliminate complexity completely.
It is to make complexity manageable.
Talk to Data Confiance about cloud assessment, migration planning and building a secure hybrid technology architecture.