From Lift and Shift to Cloud Native: Mastering Your Azure Strategy
- Sash Barige
- Aug 13, 2016
- 4 min read
Updated: Apr 1, 2024
Four years ago, if you'd told me I'd be leading our company's transformation to the cloud, I would have laughed in your face. Me? A weathered technologist who used to build end to end IT solutions for a living? Yet here I am now, eating cloud for breakfast, lunch and dinner.
It all started when our CTO came to me, eyes wild with either vision or fear (maybe both?), ranting and raving about this cloud thing that was going to disrupt our entire industry. At first, I just nodded and smiled, playing along. But he convinced me to at least investigate what possibilities the cloud might actually offer.
Well, talk about falling down a rabbit hole. The more I learned about Azure's enterprise cloud capabilities, the more I realized we were missing a huge boat. Our data center operations were bleeding money, technical debt was piling up, and we were painfully slow to launch new products. The cloud promised to solve all those problems and more.
Of course, that meant convincing risk-averse leadership that Azure was mainstream and secure enough. That Seattle company may have kick-started the cloud, but Microsoft's enterprise pedigree definitely put minds at ease. I created a spiral-bound book outlining the rationale:
Total Cost of Ownership? Check - Azure could shave millions off our DC spend. Scalability and speed? Check - we could finally free ourselves from the constraints of finite capacity. Security and Compliance? Between Microsoft's standards and rich governance controls, check and double check.
It took months of socializing, discussing, defending the strategy. But I broke through by grounding it in pragmatic steps - we'd start with a simple lift and shift, keep our options open, and then ease into cloud-native practices once we found our legs. With some small successes under our belt, people began to believe.
Now, I won't lie, it's been a long and winding road since then...
For us, the path made the most sense to start by replicating our data center footprint over to Microsoft's Azure cloud. A straightforward "lift and shift", if you will - getting us out of that captive capital investment model into a flexible, elastic cloud. Just by moving to a cloud consumption model, we could likely shave 30% off our infrastructure costs year over year.
But as the old operations saying goes, you don't want to wind up with "distributed monolithic cloud" either. Simply replicating your legacy systems and processes kind of defeats the purpose. So, in parallel to this initial lift and shift, I put together a multi-year strategy for re-architecting our applications to be truly cloud-native.
Getting the leadership team buy-in for a full Azure transformation wasn't easy in those early days. To the skeptics, I persistently made the case that going all-in on Azure solved many strategic challenges we were facing:
Azure is infinitely scalable and could support our growth projections
We could provision resources at cloud-speed instead of waiting months
We'd shift capex to OpEx, paying for only what we consumed
Built-in governance, security and compliance compared to our cobbled solutions
Ability to tap Azure's AI, analytics, IoT and other pioneering services
Here is my presentation that outlined the Azure cloud migration strategy discussed:
Slide 1: Our Path to Azure
Slide 2: The Case for Change
Constraining data center costs and technical debt
Inability to rapidly launch new digital products
Security/compliance concerns with current environment
Struggles to scale resources for fluctuating demand
Slide 3: The Azure Opportunity
Reduce infrastructure TCO by 30%+
Accelerate product cycles and time-to-market
Tap into advanced cloud services (AI, IoT, Analytics)
Robust security, DR and compliance posture
Elastic scalability and high availability
Slide 4: The Journey Ahead
Phase 1: Lift & Shift
Phase 2: Cloud-Native Transformation
Phase 3: Fully Optimized Cloud Operating Model
Slide 5: Phase 1 - Lift & Shift
Replicate on-premises footprint to IaaS (VMs, VNets, etc.)
Utilize Azure Migrate tools and services
Synchronize on-premises data
Quick ROI and bridging capability
Slide 6: Phase 2 - Cloud Native
Containerize applications with AKS
Decompose monoliths to microservices
Adopt modern data platforms like Cosmos DB
Implement DevOps processes and automation
Slide 7: Phase 3 - Optimized Cloud
Self-service deployment and GitOps
FinOps cost optimization processes
Advanced cloud architecture patterns
Unified logging, monitoring and observability
Chaos engineering and reliability practices
Slide 8: Making It Happen
Skills Transformation
Cloud architecture, DevOps, SRE (Software Reliability Engineering), FinOps
Cultural and Mindset Shift
Embrace automation, self-service, agility
Organizational Design
Product-aligned cloud platform team
Slide 9: Projected ROI
Year 1: 15% cost savings
Year 2: 25% cost savings + product velocity
Year 3: 40% TCO reduction
NPV: $25M over 5 years
Payback Period: 18 months
Slide 10: Risks & Mitigations
Technical Debt/Legacy Constraints
Fund modernization projects
Skills and Change Management
Training programs + new hires
Security and Governance
Policy-driven governance model
Slide 11: The Future is Cloud
The presentation covers the key aspects:
The drivers for cloud migration
The high-level phased migration plan
The lift & shift, re-platforming, and cloud operating goals
Projected ROI, timeline and benefits
People, process and culture shifts required
Risks and mitigations to address
It aims to provide a comprehensive overview to align stakeholders on the vision while also detailing the concrete execution plan. In my next post, I'll cover the Lift and Shift approach.
8/13/2016
Sash Barige
Links
Cloud Strategy
Phase 1
Phase 2
Phase 3
Making it Happen
DevOps Rigor
Yorumlar