Chapter 28. Why the Lift-and-Shift Model Is Unlikely to Succeed
Mike Silverman
Someone tells you to move your workloads from on premises to a cloud service provider (CSP): “It will be easy; just lift and shift.” Simple, right?
The short answer is no. A CSP and your on-premises environment are very different. Now you may be thinking, well, duh, you move to a CSP because there is some capability or feature that you can’t achieve on premises—like elasticity, resilience, global reach, or economics. All of those are good reasons to consider a move to a CSP! (Each is discussed elsewhere in this book.) But when it comes to operating workloads in CSPs versus on premises, the way they’re managed is quite different.
Consider logging as an example: the output from a CSP will likely differ from what you get from your on-prem hypervisor (or related systems). Can you commingle the logs? Unlikely. Perhaps you can map CSP events to your existing format and keep the same logging infrastructure. If you can, do you have different rules for on prem versus CSP (probably), and different automation or call routing? Does your application need to provide different event output if running in a CSP versus on prem? All things to consider, develop, test, and refine.
But that example assumes something that’s probably untrue—that the CSP can access your logging systems! Those logs ...
Become an O’Reilly member and get unlimited access to this title plus top books and audiobooks from O’Reilly and nearly 200 top publishers, thousands of courses curated by job role, 150+ live events each month,
and much more.
Read now
Unlock full access