Gloas cold DB (#8991)

Closes:

- https://github.com/sigp/lighthouse/issues/8958


  - Update the `HotColdStore` to handle storage of cold states.
- Update `BeaconSnapshot` to hold the execution envelope. This is required to make `chain_dump`-related checks sane, and will be generally useful (see: https://github.com/sigp/lighthouse/issues/8956).
- Bug fix in the `BlockReplayer` for the case where the starting state is already `Full` (we should not try to apply another payload). This happens on the cold DB path because we try to replay from the closest cached state (which is often full).
- Update `test_gloas_hot_state_hierarchy` to cover the cold DB migration.


Co-Authored-By: Michael Sproul <michael@sigmaprime.io>

Co-Authored-By: Michael Sproul <michaelsproul@users.noreply.github.com>

consensus/state_processing/src/block_replayer.rs

@@ -313,6 +313,7 @@ where
                 // indicates that the parent is full (and it hasn't already been applied).
                 state_root = if block.fork_name_unchecked().gloas_enabled()
                     && self.state.slot() == self.state.latest_block_header().slot
+                    && self.state.payload_status() == StatePayloadStatus::Pending
                 {
                     let latest_bid_block_hash = self
                         .state