In cloud migrations, the “lift-and-shift” migration approach stems from the traditional forklift datacenter migration strategy where a truck arrives after hours at location A; gremlins raid your cabinets and racks; then, depending on your local transportation union, you would be back up at location B by sunrise.
I think we can all understand why it seems appealing to employ this known strategy to your cloud migration projects, especially if your team has done a lot of data centre consolidation and transformation projects. However, there are some *gotchas* on the business and technology side to consider before jumping in.
A common refrain that we still hear today is “moving to the cloud will save money”. It’s important to know, however, exactly where those savings will come from. Equally important is knowing what changes are needed in your applications and business processes to fully realize all of the potential savings.
Through 2020, 80% of organizations conducting a “lift-and-shift” migration of internal business applications from their own data centers to public cloud IaaS will not achieve meaningful cost savings.
- Lydia Leong, Gartner (30 June 2016, ID: G00302255)
The three major business issues with “Lift-And-Shift only” Cloud Migration projects:
- No tangible operating expense savings
- Little agility gains realized, without using the cloud-native platform services
- No cost savings from decommissioning over 50% of servers(*)
(*) Yup, when doing a detailed cloud readiness assessment, our clients have reduced the number of servers in their inventories by 54% — real tangible savings.
What could Byte me?
There are technical challenges to consider as well, which we’ll cover in a future blog post, namely:
- IP address space allocation
- Understanding and managing the impact of re-IPing your servers
- Ephemeral vs Persistent Block storage
- Right-Sizing VMs to fit your IaaS provider
- Replication concerns
There are a few cases where one must lift-and-shift (Is your lease up?), and most projects we see have some element of this in them, but coupling this with as much re-platforming of high-value applications to become cloud-native as you can afford is the way to go.
Chief Migration Hacker, Tidal Migrations