Nodes that are fast-synced from snapshots will rewind to genesis if there is a rollback

[aesedepece]

I know @drcpu-github faced this issue in which nodes that are fast-synced from snapshots will rewind to genesis if there is a rollback, especially that happens right after the node completes synchronization. This needs a fix.

[drcpu-github]

I found another issue. After importing a snapshot and getting it to a stable chain, I sometimes see a UTXO consolidation error. I have only seen it twice or so over many testnets, so it's pretty hard to reproduce.

thread 'main' panicked at 'assertion failed: `(left == right)`
  left: `Some((ValueTransferOutput { pkh: PublicKeyHash { hash: [254, 162, 124, 52, 222, 228, 73, 29, 60, 44, 202, 154, 232, 97, 47, 174, 124, 122, 150, 87] }, value: 1099511627776, time_lock: 0 }, 0))`,
 right: `None`: Tried to consolidate an UTXO that was already consolidated', data_structures/src/utxo_pool.rs:96:9
note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace
[2023-05-08T20:10:00Z ERROR witnet_node::utils] Panic in ChainManager, shutting down system

This results in a node crash. Depending on how you started the node, this may result in a restart. In my case, I tend to restart them using a while true loop with the import-snapshot statement and force flag (because without it, the import statement seemed to not work) resulting in an overwrite of the database. Regardless of the restart issue, this seems like a bug worth root causing and solving.