Persistent Indexer Status in DB

[gavindoughtie]

As we will be running multiple indexers pointing at discontiguous block ranges and receiving results asynchronously, we should track status in the DB itself. As a start, we'll want to keep track of which blocks we have successfully indexed, which blocks we have tried and failed, what the ranges of contiguous blocks are, etc. This way newly launched indexers can dynamically configure themselves to avoid doing redundant work or triggering "primary key already exists" errors unnecessarily.