Physical moves are still valuable and necessary solutions, even in today increasingly virtualized and cloud based DC world.
Examples: Solaris boxes which have been running for decades or physical appliances (e.g. Ironports network and security devices).
In a nutshell: a team of engineers will unrack, wrap and load the lorry. Drive to target DCs. Unload, unwrap, rack, cable and test the devices.
What could go wrong?
- devices could get damaged during the transport (e.g. HD fail)
- the lorry might get stuck in traffic or have an accident
- cabling and connectivity in the target DC not working
How to mitigate these risks?
- review and agree a detailed schedule for the whole physical migration, involving local engineers and transportation company
- make sure the engineers have all the detailed documentation they need (racks, cabling, labels of what connects to what)
- do you have spare hardware devices in the target DC? do you have contact details of the vendor in case you need their immediate support? did you let them know about the upcoming move?
- evaluate the impact of the physical move on the whole migration – how will the rest of the migration go if the P2P fails?
If testing fails what will you do?
- Rollback (drive back to source DC) or fix forward? what is your plan B?
- How long can you (and the business) accept the downtime?