We have implemented a backup core node and scripting to switch to it if the primary node goes down or is not processing blocks. This happened in Epoch 431 when we were lucky enough to be assigned 2 blocks, and missed the first one because the core node crashed an hour before we were due to mint it. The core node automatically restarted, but then did a ledger replay and didn’t come up in time to mint the block.
If the above happens again, the backup node should take over and mint any blocks assigned while the primary node is down.