@StackStatus We are now stable and are digging into the contributing factors. The outage was ultimately caused by saturating the 20Gb network connection from our SQL Servers but the query generation differences in EF Core were the underlying cause. We're getting to the bottom of it now.
@StackStatus This query was being filtered client-side in EF Core (not SQL-side as it was with Linq2Sql). The .Where(...) was being applied client side after bringing back 52 million posts from SQL Server. This brought us offline. Still digging into why this happens. 
@StackStatus While I will dig to investigate what happened here...we're just replacing it with Dapper instead. It's drastically simpler in this case.
@StackStatus For most queries, Dapper is just a better path for us - in this case:
@StackStatus Some back-of-the-napkin math shows we'd have needed about 183Gbps of network bandwidth on the SQL Servers to stay online during that event. Nothing to be solved there, just a "fun" fact. They were dumping straight from memory and saturating the network.
@StackStatus Writing this postmortem is...weird.
"Instead of pulling back dozens of posts from the databases (filtered remotely), the StackExchange app pulled back 142 million posts per user profile load (then filtered locally)."
"Instead of pulling back dozens of posts from the databases (filtered remotely), the StackExchange app pulled back 142 million posts per user profile load (then filtered locally)."
