Super node

[i-norden]

Closed #114 and reopening here to remove some of the clutter and refocus the conversation.

Changes since that PR

• Rebased against statediff_at (to use new packages introduced there) which is rebased against staging
• Added separate command to serve up data that has already been indexed by the super node without there being an ongoing syncAndPublish process
• Fixed a flaky test
• Rename everything from seed node to super node
• Default WS port is now 8080
• Included the new db schema.sql dump
• Backfill process for periodically finding and filling in any gaps in the super node's db using an archival db with this modified geth client (VDB-934)
• Backfill unit test
• Rename misleading final field in header_cids table to uncle