Let me start with the good news: Miraheze data is transferring over much faster than we expected. In fact, everything was going swimmingly for a full migration to our new servers by Tuesday or Wednesday.
Then we had a problem where a MariaDB server was having issues dropping packets and communications and being generally unresponsive, which was the one holding metawiki and loginwiki. The Director of Site Reliability Engineering declared an emergency, ordering an immediate extraction of our assets still in the field. Which is to say, we shut down all database servers so that we could migrate as fast as possible, rather than our original plan of letting databases replicate before switching for minimal downtime.
We believe that all data is intact, and that it was a software stability problem, but at this point no one really trusts our old servers, so we're just going to move right now.
Then we had a problem where a MariaDB server was having issues dropping packets and communications and being generally unresponsive, which was the one holding metawiki and loginwiki. The Director of Site Reliability Engineering declared an emergency, ordering an immediate extraction of our assets still in the field. Which is to say, we shut down all database servers so that we could migrate as fast as possible, rather than our original plan of letting databases replicate before switching for minimal downtime.
We believe that all data is intact, and that it was a software stability problem, but at this point no one really trusts our old servers, so we're just going to move right now.
"Kitto daijoubu da yo." - Sakura Kinomoto