|
|
|
|
|
by iracic
2095 days ago
|
|
There are lot more here. Idea is to put chances on your side in some possibly long and time constrained migration.
How do you handle run that stops in the middle? How to avoid rerun of long running process that finished correctly but exited in step after? How do you revert the change to come to some previous point in time in case it is really really needed? And all that in some environment you maybe have no access but need to guide other person to act with steps defined in advance?
How do you handle if some error is fatal to stop the process and other just pass with warning printed?
Need to run some actions on each deployment?
How do you log and verify what steps were run before?
What process to use when multiple teams need to deploy to different environments and maybe different database types (different syntax, different rollback strategies)? |
|