We ultimately did not complete the infrastructure migration during this period and ended the maintenance after reverting back to our existing infrastructure.
We decided to revert the steps we took and reschedule due to a replication issue that we were not able to resolve during the maintenance window.
This infrastructure migration is moving our database and web servers from our current data center provider to Amazon Web Services (AWS).
As we move to a new environment it’s critical that we can switch back to our current environment in the event that there are any significant unforeseen issues. It would be unacceptable to have any data loss as a result of such an event, and so we have chosen to maintain a live replica of application data between the two environments for a period of several weeks after the migration.
In accordance with that we already have replication running between the two environments. That replication is using our current environment as the source and the AWS environment as the replica. Then during the migration process we reverse that and AWS becomes the source and our current environment the replica.
However, during the maintenance window we encountered replication errors when reversing the replication. These issues could not be resolved during the maintenance window. Because of that issue we called off the migration and will reschedule it for a later time after we have resolved the issue.