deployer fails during subsequent deploys due to cached snapshot endpoints

[alvrs]

When deploying to the same namespace multiple times in a short time, the deployer can't find the existing namespace, attempts to register it again, and fails because it is already registered. I think it is because we're using the getSnapshot util to find existing resources:

mud/packages/store-sync/src/getRecords.ts

Line 49 in 5ebd3d3

const logs = await getSnapshot({

But the indexer's snapshot endpoint is cached, because during a regular mud sync it's fine to catch up the remaining distance from RPC. We should update the deployer function to either catch up the remaining distance from RPC, or use the uncached SQL endpoint instead of the snapshot one.

[linear]

MUD-10 deployer fails during subsequent deploys due to cached snapshot endpoints