lighthouse/beacon_node/beacon_chain/src/fork_revert.rs

194 lines
7.2 KiB
Rust
Raw Normal View History

use crate::{BeaconForkChoiceStore, BeaconSnapshot};
use fork_choice::{ForkChoice, PayloadVerificationStatus};
use itertools::process_results;
use slog::{info, warn, Logger};
use state_processing::state_advance::complete_state_advance;
use state_processing::{
per_block_processing, per_block_processing::BlockSignatureStrategy, VerifyBlockRoot,
};
use std::sync::Arc;
v1.1.6 Fork Choice changes (#2822) ## Issue Addressed Resolves: https://github.com/sigp/lighthouse/issues/2741 Includes: https://github.com/sigp/lighthouse/pull/2853 so that we can get ssz static tests passing here on v1.1.6. If we want to merge that first, we can make this diff slightly smaller ## Proposed Changes - Changes the `justified_epoch` and `finalized_epoch` in the `ProtoArrayNode` each to an `Option<Checkpoint>`. The `Option` is necessary only for the migration, so not ideal. But does allow us to add a default logic to `None` on these fields during the database migration. - Adds a database migration from a legacy fork choice struct to the new one, search for all necessary block roots in fork choice by iterating through blocks in the db. - updates related to https://github.com/ethereum/consensus-specs/pull/2727 - We will have to update the persisted forkchoice to make sure the justified checkpoint stored is correct according to the updated fork choice logic. This boils down to setting the forkchoice store's justified checkpoint to the justified checkpoint of the block that advanced the finalized checkpoint to the current one. - AFAICT there's no migration steps necessary for the update to allow applying attestations from prior blocks, but would appreciate confirmation on that - I updated the consensus spec tests to v1.1.6 here, but they will fail until we also implement the proposer score boost updates. I confirmed that the previously failing scenario `new_finalized_slot_is_justified_checkpoint_ancestor` will now pass after the boost updates, but haven't confirmed _all_ tests will pass because I just quickly stubbed out the proposer boost test scenario formatting. - This PR now also includes proposer boosting https://github.com/ethereum/consensus-specs/pull/2730 ## Additional Info I realized checking justified and finalized roots in fork choice makes it more likely that we trigger this bug: https://github.com/ethereum/consensus-specs/pull/2727 It's possible the combination of justified checkpoint and finalized checkpoint in the forkchoice store is different from in any block in fork choice. So when trying to startup our store's justified checkpoint seems invalid to the rest of fork choice (but it should be valid). When this happens we get an `InvalidBestNode` error and fail to start up. So I'm including that bugfix in this branch. Todo: - [x] Fix fork choice tests - [x] Self review - [x] Add fix for https://github.com/ethereum/consensus-specs/pull/2727 - [x] Rebase onto Kintusgi - [x] Fix `num_active_validators` calculation as @michaelsproul pointed out - [x] Clean up db migrations Co-authored-by: realbigsean <seananderson33@gmail.com>
2021-12-13 20:43:22 +00:00
use std::time::Duration;
use store::{iter::ParentRootBlockIterator, HotColdDB, ItemStore};
use types::{BeaconState, ChainSpec, EthSpec, ForkName, Hash256, SignedBeaconBlock, Slot};
const CORRUPT_DB_MESSAGE: &str = "The database could be corrupt. Check its file permissions or \
consider deleting it by running with the --purge-db flag.";
/// Revert the head to the last block before the most recent hard fork.
///
/// This function is destructive and should only be used if there is no viable alternative. It will
/// cause the reverted blocks and states to be completely forgotten, lying dormant in the database
/// forever.
///
/// Return the `(head_block_root, head_block)` that should be used post-reversion.
pub fn revert_to_fork_boundary<E: EthSpec, Hot: ItemStore<E>, Cold: ItemStore<E>>(
current_slot: Slot,
head_block_root: Hash256,
store: Arc<HotColdDB<E, Hot, Cold>>,
spec: &ChainSpec,
log: &Logger,
) -> Result<(Hash256, SignedBeaconBlock<E>), String> {
let current_fork = spec.fork_name_at_slot::<E>(current_slot);
let fork_epoch = spec
.fork_epoch(current_fork)
.ok_or_else(|| format!("Current fork '{}' never activates", current_fork))?;
if current_fork == ForkName::Base {
return Err(format!(
"Cannot revert to before phase0 hard fork. {}",
CORRUPT_DB_MESSAGE
));
}
warn!(
log,
"Reverting invalid head block";
"target_fork" => %current_fork,
"fork_epoch" => fork_epoch,
);
let block_iter = ParentRootBlockIterator::fork_tolerant(&store, head_block_root);
process_results(block_iter, |mut iter| {
iter.find_map(|(block_root, block)| {
if block.slot() < fork_epoch.start_slot(E::slots_per_epoch()) {
Some((block_root, block))
} else {
info!(
log,
"Reverting block";
"block_root" => ?block_root,
"slot" => block.slot(),
);
None
}
})
})
.map_err(|e| {
format!(
"Error fetching blocks to revert: {:?}. {}",
e, CORRUPT_DB_MESSAGE
)
})?
.ok_or_else(|| format!("No pre-fork blocks found. {}", CORRUPT_DB_MESSAGE))
}
/// Reset fork choice to the finalized checkpoint of the supplied head state.
///
/// The supplied `head_block_root` should correspond to the most recently applied block on
/// `head_state`.
///
/// This function avoids quirks of fork choice initialization by replaying all of the blocks from
/// the checkpoint to the head.
///
/// See this issue for details: https://github.com/ethereum/consensus-specs/issues/2566
///
/// It will fail if the finalized state or any of the blocks to replay are unavailable.
///
/// WARNING: this function is destructive and causes fork choice to permanently forget all
/// chains other than the chain leading to `head_block_root`. It should only be used in extreme
/// circumstances when there is no better alternative.
pub fn reset_fork_choice_to_finalization<E: EthSpec, Hot: ItemStore<E>, Cold: ItemStore<E>>(
head_block_root: Hash256,
head_state: &BeaconState<E>,
store: Arc<HotColdDB<E, Hot, Cold>>,
spec: &ChainSpec,
) -> Result<ForkChoice<BeaconForkChoiceStore<E, Hot, Cold>, E>, String> {
// Fetch finalized block.
let finalized_checkpoint = head_state.finalized_checkpoint();
let finalized_block_root = finalized_checkpoint.root;
let finalized_block = store
.get_block(&finalized_block_root)
.map_err(|e| format!("Error loading finalized block: {:?}", e))?
.ok_or_else(|| {
format!(
"Finalized block missing for revert: {:?}",
finalized_block_root
)
})?;
// Advance finalized state to finalized epoch (to handle skipped slots).
let finalized_state_root = finalized_block.state_root();
let mut finalized_state = store
.get_state(&finalized_state_root, Some(finalized_block.slot()))
.map_err(|e| format!("Error loading finalized state: {:?}", e))?
.ok_or_else(|| {
format!(
"Finalized block state missing from database: {:?}",
finalized_state_root
)
})?;
let finalized_slot = finalized_checkpoint.epoch.start_slot(E::slots_per_epoch());
complete_state_advance(
&mut finalized_state,
Some(finalized_state_root),
finalized_slot,
spec,
)
.map_err(|e| {
format!(
"Error advancing finalized state to finalized epoch: {:?}",
e
)
})?;
let finalized_snapshot = BeaconSnapshot {
beacon_block_root: finalized_block_root,
beacon_block: finalized_block,
beacon_state: finalized_state,
};
let fc_store = BeaconForkChoiceStore::get_forkchoice_store(store.clone(), &finalized_snapshot);
let mut fork_choice = ForkChoice::from_anchor(
fc_store,
finalized_block_root,
&finalized_snapshot.beacon_block,
&finalized_snapshot.beacon_state,
)
.map_err(|e| format!("Unable to reset fork choice for revert: {:?}", e))?;
// Replay blocks from finalized checkpoint back to head.
// We do not replay attestations presently, relying on the absence of other blocks
// to guarantee `head_block_root` as the head.
let blocks = store
.load_blocks_to_replay(finalized_slot + 1, head_state.slot(), head_block_root)
.map_err(|e| format!("Error loading blocks to replay for fork choice: {:?}", e))?;
let mut state = finalized_snapshot.beacon_state;
for block in blocks {
complete_state_advance(&mut state, None, block.slot(), spec)
.map_err(|e| format!("State advance failed: {:?}", e))?;
per_block_processing(
&mut state,
&block,
None,
BlockSignatureStrategy::NoVerification,
VerifyBlockRoot::True,
spec,
)
.map_err(|e| format!("Error replaying block: {:?}", e))?;
// Setting this to unverified is the safest solution, since we don't have a way to
// retro-actively determine if they were valid or not.
//
// This scenario is so rare that it seems OK to double-verify some blocks.
let payload_verification_status = PayloadVerificationStatus::Optimistic;
let (block, _) = block.deconstruct();
fork_choice
.on_block(
block.slot(),
&block,
block.canonical_root(),
v1.1.6 Fork Choice changes (#2822) ## Issue Addressed Resolves: https://github.com/sigp/lighthouse/issues/2741 Includes: https://github.com/sigp/lighthouse/pull/2853 so that we can get ssz static tests passing here on v1.1.6. If we want to merge that first, we can make this diff slightly smaller ## Proposed Changes - Changes the `justified_epoch` and `finalized_epoch` in the `ProtoArrayNode` each to an `Option<Checkpoint>`. The `Option` is necessary only for the migration, so not ideal. But does allow us to add a default logic to `None` on these fields during the database migration. - Adds a database migration from a legacy fork choice struct to the new one, search for all necessary block roots in fork choice by iterating through blocks in the db. - updates related to https://github.com/ethereum/consensus-specs/pull/2727 - We will have to update the persisted forkchoice to make sure the justified checkpoint stored is correct according to the updated fork choice logic. This boils down to setting the forkchoice store's justified checkpoint to the justified checkpoint of the block that advanced the finalized checkpoint to the current one. - AFAICT there's no migration steps necessary for the update to allow applying attestations from prior blocks, but would appreciate confirmation on that - I updated the consensus spec tests to v1.1.6 here, but they will fail until we also implement the proposer score boost updates. I confirmed that the previously failing scenario `new_finalized_slot_is_justified_checkpoint_ancestor` will now pass after the boost updates, but haven't confirmed _all_ tests will pass because I just quickly stubbed out the proposer boost test scenario formatting. - This PR now also includes proposer boosting https://github.com/ethereum/consensus-specs/pull/2730 ## Additional Info I realized checking justified and finalized roots in fork choice makes it more likely that we trigger this bug: https://github.com/ethereum/consensus-specs/pull/2727 It's possible the combination of justified checkpoint and finalized checkpoint in the forkchoice store is different from in any block in fork choice. So when trying to startup our store's justified checkpoint seems invalid to the rest of fork choice (but it should be valid). When this happens we get an `InvalidBestNode` error and fail to start up. So I'm including that bugfix in this branch. Todo: - [x] Fix fork choice tests - [x] Self review - [x] Add fix for https://github.com/ethereum/consensus-specs/pull/2727 - [x] Rebase onto Kintusgi - [x] Fix `num_active_validators` calculation as @michaelsproul pointed out - [x] Clean up db migrations Co-authored-by: realbigsean <seananderson33@gmail.com>
2021-12-13 20:43:22 +00:00
// Reward proposer boost. We are reinforcing the canonical chain.
Duration::from_secs(0),
&state,
payload_verification_status,
spec,
)
.map_err(|e| format!("Error applying replayed block to fork choice: {:?}", e))?;
}
Ok(fork_choice)
}