DNS TTLs and the art of the zero-downtime cutover
Most migration downtime is self-inflicted and comes down to TTLs set wrong, set too late, or set on the wrong records. Here is how caching actually behaves and how to plan a cutover around it.
Most migration downtime is self-inflicted and comes down to TTLs set wrong, set too late, or set on the wrong records. Here is how caching actually behaves and how to plan a cutover around it.
Fifty sites, fifty DNS zones, hundreds of mailboxes — and one imagined catastrophe. A well-run migration is a checklist executed in a sensible order. Here is the checklist, and what to demand from the host receiving it.
Moving mailboxes between providers without losing mail, flags or folder structure. The imapsync invocations we actually use, the flags that matter, and how to sequence a cutover.