Store node's earliest available data column on disk

[eserilev]

According to the status v2 spec

When a value for `custody_group_count` is set, the `earliest_available_slot`
field in the status RPC message SHOULD reflect the slot at which the
`custody_group_count` was updated.

Currently we calculate ealiest_available_slot via beacon_chain.store.get_anchor_info().oldest_block_slot. This is problematic in situations where custody group count changes. When this happens the earliest_available_slot should be set to the slot where the cgc was changed. We need to add some additional metadata in the blobs db to keep track of the earliest_available_slot for the columns a node is required to custody.