lighthouse/beacon_node/http_api/tests/status_tests.rs

232 lines
8.3 KiB
Rust
Raw Normal View History

Implement `el_offline` and use it in the VC (#4295) ## Issue Addressed Closes https://github.com/sigp/lighthouse/issues/4291, part of #3613. ## Proposed Changes - Implement the `el_offline` field on `/eth/v1/node/syncing`. We set `el_offline=true` if: - The EL's internal status is `Offline` or `AuthFailed`, _or_ - The most recent call to `newPayload` resulted in an error (more on this in a moment). - Use the `el_offline` field in the VC to mark nodes with offline ELs as _unsynced_. These nodes will still be used, but only after synced nodes. - Overhaul the usage of `RequireSynced` so that `::No` is used almost everywhere. The `--allow-unsynced` flag was broken and had the opposite effect to intended, so it has been deprecated. - Add tests for the EL being offline on the upcheck call, and being offline due to the newPayload check. ## Why track `newPayload` errors? Tracking the EL's online/offline status is too coarse-grained to be useful in practice, because: - If the EL is timing out to some calls, it's unlikely to timeout on the `upcheck` call, which is _just_ `eth_syncing`. Every failed call is followed by an upcheck [here](https://github.com/sigp/lighthouse/blob/693886b94176faa4cb450f024696cb69cda2fe58/beacon_node/execution_layer/src/engines.rs#L372-L380), which would have the effect of masking the failure and keeping the status _online_. - The `newPayload` call is the most likely to time out. It's the call in which ELs tend to do most of their work (often 1-2 seconds), with `forkchoiceUpdated` usually returning much faster (<50ms). - If `newPayload` is failing consistently (e.g. timing out) then this is a good indication that either the node's EL is in trouble, or the network as a whole is. In the first case validator clients _should_ prefer other BNs if they have one available. In the second case, all of their BNs will likely report `el_offline` and they'll just have to proceed with trying to use them. ## Additional Changes - Add utility method `ForkName::latest` which is quite convenient for test writing, but probably other things too. - Delete some stale comments from when we used to support multiple execution nodes.
2023-05-17 05:51:56 +00:00
//! Tests related to the beacon node's sync status
use beacon_chain::{
test_utils::{AttestationStrategy, BlockStrategy, SyncCommitteeStrategy},
BlockError,
};
use eth2::StatusCode;
Implement `el_offline` and use it in the VC (#4295) ## Issue Addressed Closes https://github.com/sigp/lighthouse/issues/4291, part of #3613. ## Proposed Changes - Implement the `el_offline` field on `/eth/v1/node/syncing`. We set `el_offline=true` if: - The EL's internal status is `Offline` or `AuthFailed`, _or_ - The most recent call to `newPayload` resulted in an error (more on this in a moment). - Use the `el_offline` field in the VC to mark nodes with offline ELs as _unsynced_. These nodes will still be used, but only after synced nodes. - Overhaul the usage of `RequireSynced` so that `::No` is used almost everywhere. The `--allow-unsynced` flag was broken and had the opposite effect to intended, so it has been deprecated. - Add tests for the EL being offline on the upcheck call, and being offline due to the newPayload check. ## Why track `newPayload` errors? Tracking the EL's online/offline status is too coarse-grained to be useful in practice, because: - If the EL is timing out to some calls, it's unlikely to timeout on the `upcheck` call, which is _just_ `eth_syncing`. Every failed call is followed by an upcheck [here](https://github.com/sigp/lighthouse/blob/693886b94176faa4cb450f024696cb69cda2fe58/beacon_node/execution_layer/src/engines.rs#L372-L380), which would have the effect of masking the failure and keeping the status _online_. - The `newPayload` call is the most likely to time out. It's the call in which ELs tend to do most of their work (often 1-2 seconds), with `forkchoiceUpdated` usually returning much faster (<50ms). - If `newPayload` is failing consistently (e.g. timing out) then this is a good indication that either the node's EL is in trouble, or the network as a whole is. In the first case validator clients _should_ prefer other BNs if they have one available. In the second case, all of their BNs will likely report `el_offline` and they'll just have to proceed with trying to use them. ## Additional Changes - Add utility method `ForkName::latest` which is quite convenient for test writing, but probably other things too. - Delete some stale comments from when we used to support multiple execution nodes.
2023-05-17 05:51:56 +00:00
use execution_layer::{PayloadStatusV1, PayloadStatusV1Status};
use http_api::test_utils::InteractiveTester;
use types::{EthSpec, ExecPayload, ForkName, MinimalEthSpec, Slot};
type E = MinimalEthSpec;
/// Create a new test environment that is post-merge with `chain_depth` blocks.
async fn post_merge_tester(chain_depth: u64, validator_count: u64) -> InteractiveTester<E> {
// Test using latest fork so that we simulate conditions as similar to mainnet as possible.
2023-05-31 02:32:31 +00:00
// TODO(jimmy): We should change this back to `latest()`. These tests currently fail on Deneb because:
// 1. KZG library doesn't support Minimal spec, changing to Mainnet spec fixes some tests; BUT
// 2. `harness.process_block_result` in the test below panics due to
// `AvailabilityProcessingStatus::PendingBlobs`, and there seems to be some race
// condition going on, because the test passes if I step through the code in debug.
let mut spec = ForkName::Capella.make_genesis_spec(E::default_spec());
Implement `el_offline` and use it in the VC (#4295) ## Issue Addressed Closes https://github.com/sigp/lighthouse/issues/4291, part of #3613. ## Proposed Changes - Implement the `el_offline` field on `/eth/v1/node/syncing`. We set `el_offline=true` if: - The EL's internal status is `Offline` or `AuthFailed`, _or_ - The most recent call to `newPayload` resulted in an error (more on this in a moment). - Use the `el_offline` field in the VC to mark nodes with offline ELs as _unsynced_. These nodes will still be used, but only after synced nodes. - Overhaul the usage of `RequireSynced` so that `::No` is used almost everywhere. The `--allow-unsynced` flag was broken and had the opposite effect to intended, so it has been deprecated. - Add tests for the EL being offline on the upcheck call, and being offline due to the newPayload check. ## Why track `newPayload` errors? Tracking the EL's online/offline status is too coarse-grained to be useful in practice, because: - If the EL is timing out to some calls, it's unlikely to timeout on the `upcheck` call, which is _just_ `eth_syncing`. Every failed call is followed by an upcheck [here](https://github.com/sigp/lighthouse/blob/693886b94176faa4cb450f024696cb69cda2fe58/beacon_node/execution_layer/src/engines.rs#L372-L380), which would have the effect of masking the failure and keeping the status _online_. - The `newPayload` call is the most likely to time out. It's the call in which ELs tend to do most of their work (often 1-2 seconds), with `forkchoiceUpdated` usually returning much faster (<50ms). - If `newPayload` is failing consistently (e.g. timing out) then this is a good indication that either the node's EL is in trouble, or the network as a whole is. In the first case validator clients _should_ prefer other BNs if they have one available. In the second case, all of their BNs will likely report `el_offline` and they'll just have to proceed with trying to use them. ## Additional Changes - Add utility method `ForkName::latest` which is quite convenient for test writing, but probably other things too. - Delete some stale comments from when we used to support multiple execution nodes.
2023-05-17 05:51:56 +00:00
spec.terminal_total_difficulty = 1.into();
let tester = InteractiveTester::<E>::new(Some(spec), validator_count as usize).await;
let harness = &tester.harness;
let mock_el = harness.mock_execution_layer.as_ref().unwrap();
let execution_ctx = mock_el.server.ctx.clone();
// Move to terminal block.
mock_el.server.all_payloads_valid();
execution_ctx
.execution_block_generator
.write()
.move_to_terminal_block()
.unwrap();
// Create some chain depth.
harness.advance_slot();
harness
.extend_chain_with_sync(
chain_depth as usize,
BlockStrategy::OnCanonicalHead,
AttestationStrategy::AllValidators,
SyncCommitteeStrategy::AllValidators,
)
.await;
tester
}
/// Check `syncing` endpoint when the EL is syncing.
#[tokio::test(flavor = "multi_thread", worker_threads = 2)]
async fn el_syncing_then_synced() {
let num_blocks = E::slots_per_epoch() / 2;
let num_validators = E::slots_per_epoch();
let tester = post_merge_tester(num_blocks, num_validators).await;
let harness = &tester.harness;
let mock_el = harness.mock_execution_layer.as_ref().unwrap();
// EL syncing
mock_el.server.set_syncing_response(Ok(true));
mock_el.el.upcheck().await;
let api_response = tester.client.get_node_syncing().await.unwrap().data;
assert_eq!(api_response.el_offline, Some(false));
assert_eq!(api_response.is_optimistic, Some(false));
assert_eq!(api_response.is_syncing, false);
// EL synced
mock_el.server.set_syncing_response(Ok(false));
mock_el.el.upcheck().await;
let api_response = tester.client.get_node_syncing().await.unwrap().data;
assert_eq!(api_response.el_offline, Some(false));
assert_eq!(api_response.is_optimistic, Some(false));
assert_eq!(api_response.is_syncing, false);
}
/// Check `syncing` endpoint when the EL is offline (errors on upcheck).
#[tokio::test(flavor = "multi_thread", worker_threads = 2)]
async fn el_offline() {
let num_blocks = E::slots_per_epoch() / 2;
let num_validators = E::slots_per_epoch();
let tester = post_merge_tester(num_blocks, num_validators).await;
let harness = &tester.harness;
let mock_el = harness.mock_execution_layer.as_ref().unwrap();
// EL offline
mock_el.server.set_syncing_response(Err("offline".into()));
mock_el.el.upcheck().await;
let api_response = tester.client.get_node_syncing().await.unwrap().data;
assert_eq!(api_response.el_offline, Some(true));
assert_eq!(api_response.is_optimistic, Some(false));
assert_eq!(api_response.is_syncing, false);
}
/// Check `syncing` endpoint when the EL errors on newPaylod but is not fully offline.
#[tokio::test(flavor = "multi_thread", worker_threads = 2)]
async fn el_error_on_new_payload() {
let num_blocks = E::slots_per_epoch() / 2;
let num_validators = E::slots_per_epoch();
let tester = post_merge_tester(num_blocks, num_validators).await;
let harness = &tester.harness;
let mock_el = harness.mock_execution_layer.as_ref().unwrap();
// Make a block.
let pre_state = harness.get_current_state();
2023-05-30 13:31:09 +00:00
let (block_contents, _) = harness
Implement `el_offline` and use it in the VC (#4295) ## Issue Addressed Closes https://github.com/sigp/lighthouse/issues/4291, part of #3613. ## Proposed Changes - Implement the `el_offline` field on `/eth/v1/node/syncing`. We set `el_offline=true` if: - The EL's internal status is `Offline` or `AuthFailed`, _or_ - The most recent call to `newPayload` resulted in an error (more on this in a moment). - Use the `el_offline` field in the VC to mark nodes with offline ELs as _unsynced_. These nodes will still be used, but only after synced nodes. - Overhaul the usage of `RequireSynced` so that `::No` is used almost everywhere. The `--allow-unsynced` flag was broken and had the opposite effect to intended, so it has been deprecated. - Add tests for the EL being offline on the upcheck call, and being offline due to the newPayload check. ## Why track `newPayload` errors? Tracking the EL's online/offline status is too coarse-grained to be useful in practice, because: - If the EL is timing out to some calls, it's unlikely to timeout on the `upcheck` call, which is _just_ `eth_syncing`. Every failed call is followed by an upcheck [here](https://github.com/sigp/lighthouse/blob/693886b94176faa4cb450f024696cb69cda2fe58/beacon_node/execution_layer/src/engines.rs#L372-L380), which would have the effect of masking the failure and keeping the status _online_. - The `newPayload` call is the most likely to time out. It's the call in which ELs tend to do most of their work (often 1-2 seconds), with `forkchoiceUpdated` usually returning much faster (<50ms). - If `newPayload` is failing consistently (e.g. timing out) then this is a good indication that either the node's EL is in trouble, or the network as a whole is. In the first case validator clients _should_ prefer other BNs if they have one available. In the second case, all of their BNs will likely report `el_offline` and they'll just have to proceed with trying to use them. ## Additional Changes - Add utility method `ForkName::latest` which is quite convenient for test writing, but probably other things too. - Delete some stale comments from when we used to support multiple execution nodes.
2023-05-17 05:51:56 +00:00
.make_block(pre_state, Slot::new(num_blocks + 1))
.await;
2023-05-30 13:31:09 +00:00
let block = block_contents.0;
Implement `el_offline` and use it in the VC (#4295) ## Issue Addressed Closes https://github.com/sigp/lighthouse/issues/4291, part of #3613. ## Proposed Changes - Implement the `el_offline` field on `/eth/v1/node/syncing`. We set `el_offline=true` if: - The EL's internal status is `Offline` or `AuthFailed`, _or_ - The most recent call to `newPayload` resulted in an error (more on this in a moment). - Use the `el_offline` field in the VC to mark nodes with offline ELs as _unsynced_. These nodes will still be used, but only after synced nodes. - Overhaul the usage of `RequireSynced` so that `::No` is used almost everywhere. The `--allow-unsynced` flag was broken and had the opposite effect to intended, so it has been deprecated. - Add tests for the EL being offline on the upcheck call, and being offline due to the newPayload check. ## Why track `newPayload` errors? Tracking the EL's online/offline status is too coarse-grained to be useful in practice, because: - If the EL is timing out to some calls, it's unlikely to timeout on the `upcheck` call, which is _just_ `eth_syncing`. Every failed call is followed by an upcheck [here](https://github.com/sigp/lighthouse/blob/693886b94176faa4cb450f024696cb69cda2fe58/beacon_node/execution_layer/src/engines.rs#L372-L380), which would have the effect of masking the failure and keeping the status _online_. - The `newPayload` call is the most likely to time out. It's the call in which ELs tend to do most of their work (often 1-2 seconds), with `forkchoiceUpdated` usually returning much faster (<50ms). - If `newPayload` is failing consistently (e.g. timing out) then this is a good indication that either the node's EL is in trouble, or the network as a whole is. In the first case validator clients _should_ prefer other BNs if they have one available. In the second case, all of their BNs will likely report `el_offline` and they'll just have to proceed with trying to use them. ## Additional Changes - Add utility method `ForkName::latest` which is quite convenient for test writing, but probably other things too. - Delete some stale comments from when we used to support multiple execution nodes.
2023-05-17 05:51:56 +00:00
let block_hash = block
.message()
.body()
.execution_payload()
.unwrap()
.block_hash();
// Make sure `newPayload` errors for the new block.
mock_el
.server
.set_new_payload_error(block_hash, "error".into());
// Attempt to process the block, which should error.
harness.advance_slot();
assert!(matches!(
harness.process_block_result(block.clone()).await,
Err(BlockError::ExecutionPayloadError(_))
));
// The EL should now be *offline* according to the API.
let api_response = tester.client.get_node_syncing().await.unwrap().data;
assert_eq!(api_response.el_offline, Some(true));
assert_eq!(api_response.is_optimistic, Some(false));
assert_eq!(api_response.is_syncing, false);
// Processing a block successfully should remove the status.
mock_el.server.set_new_payload_status(
block_hash,
PayloadStatusV1 {
status: PayloadStatusV1Status::Valid,
latest_valid_hash: Some(block_hash),
validation_error: None,
},
);
harness.process_block_result(block).await.unwrap();
let api_response = tester.client.get_node_syncing().await.unwrap().data;
assert_eq!(api_response.el_offline, Some(false));
assert_eq!(api_response.is_optimistic, Some(false));
assert_eq!(api_response.is_syncing, false);
}
/// Check `node health` endpoint when the EL is offline.
#[tokio::test(flavor = "multi_thread", worker_threads = 2)]
async fn node_health_el_offline() {
let num_blocks = E::slots_per_epoch() / 2;
let num_validators = E::slots_per_epoch();
let tester = post_merge_tester(num_blocks, num_validators).await;
let harness = &tester.harness;
let mock_el = harness.mock_execution_layer.as_ref().unwrap();
// EL offline
mock_el.server.set_syncing_response(Err("offline".into()));
mock_el.el.upcheck().await;
let status = tester.client.get_node_health().await;
match status {
Ok(_) => {
panic!("should return 503 error status code");
}
Err(e) => {
assert_eq!(e.status().unwrap(), 503);
}
}
}
/// Check `node health` endpoint when the EL is online and synced.
#[tokio::test(flavor = "multi_thread", worker_threads = 2)]
async fn node_health_el_online_and_synced() {
let num_blocks = E::slots_per_epoch() / 2;
let num_validators = E::slots_per_epoch();
let tester = post_merge_tester(num_blocks, num_validators).await;
let harness = &tester.harness;
let mock_el = harness.mock_execution_layer.as_ref().unwrap();
// EL synced
mock_el.server.set_syncing_response(Ok(false));
mock_el.el.upcheck().await;
let status = tester.client.get_node_health().await;
match status {
Ok(response) => {
assert_eq!(response, StatusCode::OK);
}
Err(_) => {
panic!("should return 200 status code");
}
}
}
/// Check `node health` endpoint when the EL is online but not synced.
#[tokio::test(flavor = "multi_thread", worker_threads = 2)]
async fn node_health_el_online_and_not_synced() {
let num_blocks = E::slots_per_epoch() / 2;
let num_validators = E::slots_per_epoch();
let tester = post_merge_tester(num_blocks, num_validators).await;
let harness = &tester.harness;
let mock_el = harness.mock_execution_layer.as_ref().unwrap();
// EL not synced
harness.advance_slot();
mock_el.server.all_payloads_syncing(true);
harness
.extend_chain(
1,
BlockStrategy::OnCanonicalHead,
AttestationStrategy::AllValidators,
)
.await;
let status = tester.client.get_node_health().await;
match status {
Ok(response) => {
assert_eq!(response, StatusCode::PARTIAL_CONTENT);
}
Err(_) => {
panic!("should return 206 status code");
}
}
}