Well, it all comes down to your environment and how it can handle down-time. Personally I would migrate if possible due to the complexity of the domain rename process.
What you need to understand is that the domain rename process is complex and in some ways "uncontrolled". Meaning that if you start it, it will try to finish or fail.
Things to understand:
- Headless Management: Rendom will not use AD Repl, each DC will be contacted individually.
- Forest will be "offline" during the process: The time is proportional to the number of DCs.
- DC is either successful or must be removed from the forest.
- Member computers must be rebooted twise after all DCs are updated, note that legacy OS needs to unjoin/join the domain.
- DNS host names are not automatically changed during rendom: Primary DNS suffix of the DC will not mathc the new domain DNS name. Requires additional steps after rendom.
- DNS suffix on member computers will not match for a period of time: Time it takes is proportional to the number of machines in the domain and if it is auto updated or not.
- Runs from a separate computer
- CLI interface
- Each DC is changed independently
- Step-by-step, steps must succeed on every DC or it stops
- Steps are idempotent
3 DC states:
- Initial
- Prepared
- Final (success or fail)
Note that Prepared to Final can only be accomplished if every DC in the forest has reached the required state!
Overview of the steps:
- Manually specify the new forest structure
- Generate instructions encoded as script
- Transfer the script to each DC
- Verify the script on every DC to see if it is ready to execute the instructions
- Execute instructions on each DC (forest offline during this step)
- Fix-up Policy metadata
- Clean-up metadata written to the directory
Remember that this is just a basic overview of how it works, you must test, test again and test again in a lab before trying it in production!
Wednesday, June 12, 2013
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment