gnosis/lighthouse
1
0
Fork 0
mirror of https://github.com/sigp/lighthouse.git synced 2026-03-15 02:42:38 +00:00
Code Issues Packages Projects Releases Wiki Activity

Use async code when interacting with EL (#3244)

Browse Source 
## Overview

This rather extensive PR achieves two primary goals:

1. Uses the finalized/justified checkpoints of fork choice (FC), rather than that of the head state.
2. Refactors fork choice, block production and block processing to `async` functions.

Additionally, it achieves:

- Concurrent forkchoice updates to the EL and cache pruning after a new head is selected.
- Concurrent "block packing" (attestations, etc) and execution payload retrieval during block production.
- Concurrent per-block-processing and execution payload verification during block processing.
- The `Arc`-ification of `SignedBeaconBlock` during block processing (it's never mutated, so why not?):
    - I had to do this to deal with sending blocks into spawned tasks.
    - Previously we were cloning the beacon block at least 2 times during each block processing, these clones are either removed or turned into cheaper `Arc` clones.
    - We were also `Box`-ing and un-`Box`-ing beacon blocks as they moved throughout the networking crate. This is not a big deal, but it's nice to avoid shifting things between the stack and heap.
    - Avoids cloning *all the blocks* in *every chain segment* during sync.
    - It also has the potential to clean up our code where we need to pass an *owned* block around so we can send it back in the case of an error (I didn't do much of this, my PR is already big enough 😅)
- The `BeaconChain::HeadSafetyStatus` struct was removed. It was an old relic from prior merge specs.

For motivation for this change, see https://github.com/sigp/lighthouse/pull/3244#issuecomment-1160963273

## Changes to `canonical_head` and `fork_choice`

Previously, the `BeaconChain` had two separate fields:

```
canonical_head: RwLock<Snapshot>,
fork_choice: RwLock<BeaconForkChoice>
```

Now, we have grouped these values under a single struct:

```
canonical_head: CanonicalHead {
  cached_head: RwLock<Arc<Snapshot>>,
  fork_choice: RwLock<BeaconForkChoice>
} 
```

Apart from ergonomics, the only *actual* change here is wrapping the canonical head snapshot in an `Arc`. This means that we no longer need to hold the `cached_head` (`canonical_head`, in old terms) lock when we want to pull some values from it. This was done to avoid deadlock risks by preventing functions from acquiring (and holding) the `cached_head` and `fork_choice` locks simultaneously.

## Breaking Changes

### The `state` (root) field in the `finalized_checkpoint` SSE event

Consider the scenario where epoch `n` is just finalized, but `start_slot(n)` is skipped. There are two state roots we might in the `finalized_checkpoint` SSE event:

1. The state root of the finalized block, which is `get_block(finalized_checkpoint.root).state_root`.
4. The state root at slot of `start_slot(n)`, which would be the state from (1), but "skipped forward" through any skip slots.

Previously, Lighthouse would choose (2). However, we can see that when [Teku generates that event](de2b2801c8/data/beaconrestapi/src/main/java/tech/pegasys/teku/beaconrestapi/handlers/v1/events/EventSubscriptionManager.java (L171-L182)) it uses [`getStateRootFromBlockRoot`](de2b2801c8/data/provider/src/main/java/tech/pegasys/teku/api/ChainDataProvider.java (L336-L341)) which uses (1).

I have switched Lighthouse from (2) to (1). I think it's a somewhat arbitrary choice between the two, where (1) is easier to compute and is consistent with Teku.

## Notes for Reviewers

I've renamed `BeaconChain::fork_choice` to `BeaconChain::recompute_head`. Doing this helped ensure I broke all previous uses of fork choice and I also find it more descriptive. It describes an action and can't be confused with trying to get a reference to the `ForkChoice` struct.

I've changed the ordering of SSE events when a block is received. It used to be `[block, finalized, head]` and now it's `[block, head, finalized]`. It was easier this way and I don't think we were making any promises about SSE event ordering so it's not "breaking".

I've made it so fork choice will run when it's first constructed. I did this because I wanted to have a cached version of the last call to `get_head`. Ensuring `get_head` has been run *at least once* means that the cached values doesn't need to wrapped in an `Option`. This was fairly simple, it just involved passing a `slot` to the constructor so it knows *when* it's being run. When loading a fork choice from the store and a slot clock isn't handy I've just used the `slot` that was saved in the `fork_choice_store`. That seems like it would be a faithful representation of the slot when we saved it.

I added the `genesis_time: u64` to the `BeaconChain`. It's small, constant and nice to have around.

Since we're using FC for the fin/just checkpoints, we no longer get the `0x00..00` roots at genesis. You can see I had to remove a work-around in `ef-tests` here: b56be3bc2. I can't find any reason why this would be an issue, if anything I think it'll be better since the genesis-alias has caught us out a few times (0x00..00 isn't actually a real root). Edit: I did find a case where the `network` expected the 0x00..00 alias and patched it here: 3f26ac3e2.

You'll notice a lot of changes in tests. Generally, tests should be functionally equivalent. Here are the things creating the most diff-noise in tests:
- Changing tests to be `tokio::async` tests.
- Adding `.await` to fork choice, block processing and block production functions.
- Refactor of the `canonical_head` "API" provided by the `BeaconChain`. E.g., `chain.canonical_head.cached_head()` instead of `chain.canonical_head.read()`.
- Wrapping `SignedBeaconBlock` in an `Arc`.
- In the `beacon_chain/tests/block_verification`, we can't use the `lazy_static` `CHAIN_SEGMENT` variable anymore since it's generated with an async function. We just generate it in each test, not so efficient but hopefully insignificant.

I had to disable `rayon` concurrent tests in the `fork_choice` tests. This is because the use of `rayon` and `block_on` was causing a panic.

Co-authored-by: Mac L <mjladson@pm.me>
This commit is contained in:
Paul Hauner
2022-07-03 05:36:50 +00:00
parent e5212f1320
commit be4e261e74
106 changed files with 6515 additions and 4538 deletions
Download Patch File Download Diff File Expand all files Collapse all files

219
beacon_node/http_api/tests/tests.rs
View File

@@ -11,7 +11,6 @@ use eth2::{
types::*,
BeaconNodeHttpClient, Error, StatusCode, Timeouts,
};
use execution_layer::test_utils::MockExecutionLayer;
use futures::stream::{Stream, StreamExt};
use futures::FutureExt;
use lighthouse_network::{Enr, EnrExt, PeerId};
@@ -21,7 +20,6 @@ use slot_clock::SlotClock;
use state_processing::per_slot_processing;
use std::convert::TryInto;
use std::sync::Arc;
use task_executor::test_utils::TestRuntime;
use tokio::sync::{mpsc, oneshot};
use tokio::time::Duration;
use tree_hash::TreeHash;
@@ -52,6 +50,7 @@ const SKIPPED_SLOTS: &[u64] = &[
];
struct ApiTester {
harness: Arc<BeaconChainHarness<EphemeralHarnessType<E>>>,
chain: Arc<BeaconChain<EphemeralHarnessType<E>>>,
client: BeaconNodeHttpClient,
next_block: SignedBeaconBlock<E>,
@@ -62,14 +61,9 @@ struct ApiTester {
proposer_slashing: ProposerSlashing,
voluntary_exit: SignedVoluntaryExit,
_server_shutdown: oneshot::Sender<()>,
validator_keypairs: Vec<Keypair>,
network_rx: mpsc::UnboundedReceiver<NetworkMessage<E>>,
local_enr: Enr,
external_peer_id: PeerId,
// This is never directly accessed, but adding it creates a payload cache, which we use in tests here.
#[allow(dead_code)]
mock_el: Option<MockExecutionLayer<E>>,
_runtime: TestRuntime,
}
impl ApiTester {
@@ -81,12 +75,14 @@ impl ApiTester {
}
pub async fn new_from_spec(spec: ChainSpec) -> Self {
let harness = BeaconChainHarness::builder(MainnetEthSpec)
.spec(spec.clone())
.deterministic_keypairs(VALIDATOR_COUNT)
.fresh_ephemeral_store()
.mock_execution_layer()
.build();
let harness = Arc::new(
BeaconChainHarness::builder(MainnetEthSpec)
.spec(spec.clone())
.deterministic_keypairs(VALIDATOR_COUNT)
.fresh_ephemeral_store()
.mock_execution_layer()
.build(),
);
harness.advance_slot();
@@ -94,17 +90,19 @@ impl ApiTester {
let slot = harness.chain.slot().unwrap().as_u64();
if !SKIPPED_SLOTS.contains(&slot) {
harness.extend_chain(
1,
BlockStrategy::OnCanonicalHead,
AttestationStrategy::AllValidators,
);
harness
.extend_chain(
1,
BlockStrategy::OnCanonicalHead,
AttestationStrategy::AllValidators,
)
.await;
}
harness.advance_slot();
}
let head = harness.chain.head().unwrap();
let head = harness.chain.head_snapshot();
assert_eq!(
harness.chain.slot().unwrap(),
@@ -112,12 +110,14 @@ impl ApiTester {
"precondition: current slot is one after head"
);
let (next_block, _next_state) =
harness.make_block(head.beacon_state.clone(), harness.chain.slot().unwrap());
let (next_block, _next_state) = harness
.make_block(head.beacon_state.clone(), harness.chain.slot().unwrap())
.await;
// `make_block` adds random graffiti, so this will produce an alternate block
let (reorg_block, _reorg_state) =
harness.make_block(head.beacon_state.clone(), harness.chain.slot().unwrap());
let (reorg_block, _reorg_state) = harness
.make_block(head.beacon_state.clone(), harness.chain.slot().unwrap())
.await;
let head_state_root = head.beacon_state_root();
let attestations = harness
@@ -168,15 +168,19 @@ impl ApiTester {
let chain = harness.chain.clone();
assert_eq!(
chain.head_info().unwrap().finalized_checkpoint.epoch,
chain
.canonical_head
.cached_head()
.finalized_checkpoint()
.epoch,
2,
"precondition: finality"
);
assert_eq!(
chain
.head_info()
.unwrap()
.current_justified_checkpoint
.canonical_head
.cached_head()
.justified_checkpoint()
.epoch,
3,
"precondition: justification"
@@ -206,6 +210,7 @@ impl ApiTester {
);
Self {
harness,
chain,
client,
next_block,
@@ -216,32 +221,33 @@ impl ApiTester {
proposer_slashing,
voluntary_exit,
_server_shutdown: shutdown_tx,
validator_keypairs: harness.validator_keypairs,
network_rx,
local_enr,
external_peer_id,
mock_el: harness.mock_execution_layer,
_runtime: harness.runtime,
}
}
pub async fn new_from_genesis() -> Self {
let harness = BeaconChainHarness::builder(MainnetEthSpec)
.default_spec()
.deterministic_keypairs(VALIDATOR_COUNT)
.fresh_ephemeral_store()
.build();
let harness = Arc::new(
BeaconChainHarness::builder(MainnetEthSpec)
.default_spec()
.deterministic_keypairs(VALIDATOR_COUNT)
.fresh_ephemeral_store()
.build(),
);
harness.advance_slot();
let head = harness.chain.head().unwrap();
let head = harness.chain.head_snapshot();
let (next_block, _next_state) =
harness.make_block(head.beacon_state.clone(), harness.chain.slot().unwrap());
let (next_block, _next_state) = harness
.make_block(head.beacon_state.clone(), harness.chain.slot().unwrap())
.await;
// `make_block` adds random graffiti, so this will produce an alternate block
let (reorg_block, _reorg_state) =
harness.make_block(head.beacon_state.clone(), harness.chain.slot().unwrap());
let (reorg_block, _reorg_state) = harness
.make_block(head.beacon_state.clone(), harness.chain.slot().unwrap())
.await;
let head_state_root = head.beacon_state_root();
let attestations = harness
@@ -286,6 +292,7 @@ impl ApiTester {
);
Self {
harness,
chain,
client,
next_block,
@@ -296,15 +303,16 @@ impl ApiTester {
proposer_slashing,
voluntary_exit,
_server_shutdown: shutdown_tx,
validator_keypairs: harness.validator_keypairs,
network_rx,
local_enr,
external_peer_id,
mock_el: None,
_runtime: harness.runtime,
}
}
fn validator_keypairs(&self) -> &[Keypair] {
&self.harness.validator_keypairs
}
fn skip_slots(self, count: u64) -> Self {
for _ in 0..count {
self.chain
@@ -329,7 +337,9 @@ impl ApiTester {
StateId::Slot(Slot::from(SKIPPED_SLOTS[3])),
StateId::Root(Hash256::zero()),
];
ids.push(StateId::Root(self.chain.head_info().unwrap().state_root));
ids.push(StateId::Root(
self.chain.canonical_head.cached_head().head_state_root(),
));
ids
}
@@ -347,13 +357,20 @@ impl ApiTester {
BlockId::Slot(Slot::from(SKIPPED_SLOTS[3])),
BlockId::Root(Hash256::zero()),
];
ids.push(BlockId::Root(self.chain.head_info().unwrap().block_root));
ids.push(BlockId::Root(
self.chain.canonical_head.cached_head().head_block_root(),
));
ids
}
fn get_state(&self, state_id: StateId) -> Option<BeaconState<E>> {
match state_id {
StateId::Head => Some(self.chain.head().unwrap().beacon_state),
StateId::Head => Some(
self.chain
.head_snapshot()
.beacon_state
.clone_with_only_committee_caches(),
),
StateId::Genesis => self
.chain
.get_state(&self.chain.genesis_state_root, None)
@@ -361,9 +378,9 @@ impl ApiTester {
StateId::Finalized => {
let finalized_slot = self
.chain
.head_info()
.unwrap()
.finalized_checkpoint
.canonical_head
.cached_head()
.finalized_checkpoint()
.epoch
.start_slot(E::slots_per_epoch());
@@ -378,9 +395,9 @@ impl ApiTester {
StateId::Justified => {
let justified_slot = self
.chain
.head_info()
.unwrap()
.current_justified_checkpoint
.canonical_head
.cached_head()
.justified_checkpoint()
.epoch
.start_slot(E::slots_per_epoch());
@@ -404,7 +421,7 @@ impl ApiTester {
pub async fn test_beacon_genesis(self) -> Self {
let result = self.client.get_beacon_genesis().await.unwrap().data;
let state = self.chain.head().unwrap().beacon_state;
let state = &self.chain.head_snapshot().beacon_state;
let expected = GenesisData {
genesis_time: state.genesis_time(),
genesis_validators_root: state.genesis_validators_root(),
@@ -426,14 +443,14 @@ impl ApiTester {
.map(|res| res.data.root);
let expected = match state_id {
StateId::Head => Some(self.chain.head_info().unwrap().state_root),
StateId::Head => Some(self.chain.canonical_head.cached_head().head_state_root()),
StateId::Genesis => Some(self.chain.genesis_state_root),
StateId::Finalized => {
let finalized_slot = self
.chain
.head_info()
.unwrap()
.finalized_checkpoint
.canonical_head
.cached_head()
.finalized_checkpoint()
.epoch
.start_slot(E::slots_per_epoch());
@@ -442,9 +459,9 @@ impl ApiTester {
StateId::Justified => {
let justified_slot = self
.chain
.head_info()
.unwrap()
.current_justified_checkpoint
.canonical_head
.cached_head()
.justified_checkpoint()
.epoch
.start_slot(E::slots_per_epoch());
@@ -754,14 +771,20 @@ impl ApiTester {
fn get_block_root(&self, block_id: BlockId) -> Option<Hash256> {
match block_id {
BlockId::Head => Some(self.chain.head_info().unwrap().block_root),
BlockId::Head => Some(self.chain.canonical_head.cached_head().head_block_root()),
BlockId::Genesis => Some(self.chain.genesis_block_root),
BlockId::Finalized => Some(self.chain.head_info().unwrap().finalized_checkpoint.root),
BlockId::Finalized => Some(
self.chain
.canonical_head
.cached_head()
.finalized_checkpoint()
.root,
),
BlockId::Justified => Some(
self.chain
.head_info()
.unwrap()
.current_justified_checkpoint
.canonical_head
.cached_head()
.justified_checkpoint()
.root,
),
BlockId::Slot(slot) => self
@@ -1322,7 +1345,7 @@ impl ApiTester {
pub async fn test_get_node_syncing(self) -> Self {
let result = self.client.get_node_syncing().await.unwrap().data;
let head_slot = self.chain.head_info().unwrap().slot;
let head_slot = self.chain.canonical_head.cached_head().head_slot();
let sync_distance = self.chain.slot().unwrap() - head_slot;
let expected = SyncingData {
@@ -1536,7 +1559,7 @@ impl ApiTester {
}
fn validator_count(&self) -> usize {
self.chain.head().unwrap().beacon_state.validators().len()
self.chain.head_snapshot().beacon_state.validators().len()
}
fn interesting_validator_indices(&self) -> Vec<Vec<u64>> {
@@ -1621,7 +1644,7 @@ impl ApiTester {
WhenSlotSkipped::Prev,
)
.unwrap()
.unwrap_or(self.chain.head_beacon_block_root().unwrap());
.unwrap_or(self.chain.head_beacon_block_root());
assert_eq!(results.dependent_root, dependent_root);
@@ -1696,7 +1719,7 @@ impl ApiTester {
WhenSlotSkipped::Prev,
)
.unwrap()
.unwrap_or(self.chain.head_beacon_block_root().unwrap());
.unwrap_or(self.chain.head_beacon_block_root());
// Presently, the beacon chain harness never runs the code that primes the proposer
// cache. If this changes in the future then we'll need some smarter logic here, but
@@ -1824,7 +1847,7 @@ impl ApiTester {
WhenSlotSkipped::Prev,
)
.unwrap()
.unwrap_or(self.chain.head_beacon_block_root().unwrap());
.unwrap_or(self.chain.head_beacon_block_root());
self.client
.get_validator_duties_proposer(current_epoch)
@@ -1878,7 +1901,7 @@ impl ApiTester {
}
pub async fn test_block_production(self) -> Self {
let fork = self.chain.head_info().unwrap().fork;
let fork = self.chain.canonical_head.cached_head().head_fork();
let genesis_validators_root = self.chain.genesis_validators_root;
for _ in 0..E::slots_per_epoch() * 3 {
@@ -1898,7 +1921,7 @@ impl ApiTester {
let proposer_pubkey = (&proposer_pubkey_bytes).try_into().unwrap();
let sk = self
.validator_keypairs
.validator_keypairs()
.iter()
.find(|kp| kp.pk == proposer_pubkey)
.map(|kp| kp.sk.clone())
@@ -1926,7 +1949,7 @@ impl ApiTester {
self.client.post_beacon_blocks(&signed_block).await.unwrap();
assert_eq!(self.chain.head_beacon_block().unwrap(), signed_block);
assert_eq!(self.chain.head_beacon_block().as_ref(), &signed_block);
self.chain.slot_clock.set_slot(slot.as_u64() + 1);
}
@@ -1957,7 +1980,7 @@ impl ApiTester {
}
pub async fn test_block_production_verify_randao_invalid(self) -> Self {
let fork = self.chain.head_info().unwrap().fork;
let fork = self.chain.canonical_head.cached_head().head_fork();
let genesis_validators_root = self.chain.genesis_validators_root;
for _ in 0..E::slots_per_epoch() {
@@ -1977,7 +2000,7 @@ impl ApiTester {
let proposer_pubkey = (&proposer_pubkey_bytes).try_into().unwrap();
let sk = self
.validator_keypairs
.validator_keypairs()
.iter()
.find(|kp| kp.pk == proposer_pubkey)
.map(|kp| kp.sk.clone())
@@ -2040,7 +2063,7 @@ impl ApiTester {
}
pub async fn test_get_validator_attestation_data(self) -> Self {
let mut state = self.chain.head_beacon_state().unwrap();
let mut state = self.chain.head_beacon_state_cloned();
let slot = state.slot();
state
.build_committee_cache(RelativeEpoch::Current, &self.chain.spec)
@@ -2070,7 +2093,6 @@ impl ApiTester {
let attestation = self
.chain
.head_beacon_block()
.unwrap()
.message()
.body()
.attestations()[0]
@@ -2098,7 +2120,7 @@ impl ApiTester {
let slot = self.chain.slot().unwrap();
let epoch = self.chain.epoch().unwrap();
let mut head = self.chain.head().unwrap();
let mut head = self.chain.head_snapshot().as_ref().clone();
while head.beacon_state.current_epoch() < epoch {
per_slot_processing(&mut head.beacon_state, None, &self.chain.spec).unwrap();
}
@@ -2114,7 +2136,7 @@ impl ApiTester {
.client
.post_validator_duties_attester(
epoch,
(0..self.validator_keypairs.len() as u64)
(0..self.validator_keypairs().len() as u64)
.collect::<Vec<u64>>()
.as_slice(),
)
@@ -2123,7 +2145,7 @@ impl ApiTester {
.data;
let (i, kp, duty, proof) = self
.validator_keypairs
.validator_keypairs()
.iter()
.enumerate()
.find_map(|(i, kp)| {
@@ -2238,9 +2260,9 @@ impl ApiTester {
let mut registrations = vec![];
let mut fee_recipients = vec![];
let fork = self.chain.head().unwrap().beacon_state.fork();
let fork = self.chain.head_snapshot().beacon_state.fork();
for (val_index, keypair) in self.validator_keypairs.iter().enumerate() {
for (val_index, keypair) in self.validator_keypairs().iter().enumerate() {
let pubkey = keypair.pk.compress();
let fee_recipient = Address::from_low_u64_be(val_index as u64);
@@ -2273,8 +2295,7 @@ impl ApiTester {
for (val_index, (_, fee_recipient)) in self
.chain
.head()
.unwrap()
.head_snapshot()
.beacon_state
.validators()
.into_iter()
@@ -2416,7 +2437,7 @@ impl ApiTester {
pub async fn test_post_lighthouse_liveness(self) -> Self {
let epoch = self.chain.epoch().unwrap();
let head_state = self.chain.head_beacon_state().unwrap();
let head_state = self.chain.head_beacon_state_cloned();
let indices = (0..head_state.validators().len())
.map(|i| i as u64)
.collect::<Vec<_>>();
@@ -2533,7 +2554,7 @@ impl ApiTester {
let block_root = self.next_block.canonical_root();
// current_duty_dependent_root = block root because this is the first slot of the epoch
let current_duty_dependent_root = self.chain.head_beacon_block_root().unwrap();
let current_duty_dependent_root = self.chain.head_beacon_block_root();
let current_slot = self.chain.slot().unwrap();
let next_slot = self.next_block.slot();
let finalization_distance = E::slots_per_epoch() * 2;
@@ -2556,17 +2577,21 @@ impl ApiTester {
epoch_transition: true,
});
let finalized_block_root = self
.chain
.block_root_at_slot(next_slot - finalization_distance, WhenSlotSkipped::Prev)
.unwrap()
.unwrap();
let finalized_block = self
.chain
.get_blinded_block(&finalized_block_root)
.unwrap()
.unwrap();
let finalized_state_root = finalized_block.state_root();
let expected_finalized = EventKind::FinalizedCheckpoint(SseFinalizedCheckpoint {
block: self
.chain
.block_root_at_slot(next_slot - finalization_distance, WhenSlotSkipped::Prev)
.unwrap()
.unwrap(),
state: self
.chain
.state_root_at_slot(next_slot - finalization_distance)
.unwrap()
.unwrap(),
block: finalized_block_root,
state: finalized_state_root,
epoch: Epoch::new(3),
});
@@ -2578,7 +2603,7 @@ impl ApiTester {
let block_events = poll_events(&mut events_future, 3, Duration::from_millis(10000)).await;
assert_eq!(
block_events.as_slice(),
&[expected_block, expected_finalized, expected_head]
&[expected_block, expected_head, expected_finalized]
);
// Test a reorg event