astar_collator/parachain/

service.rs

// This file is part of Astar.

// Copyright (C) Stake Technologies Pte.Ltd.
// SPDX-License-Identifier: GPL-3.0-or-later

// Astar is free software: you can redistribute it and/or modify
// it under the terms of the GNU General Public License as published by
// the Free Software Foundation, either version 3 of the License, or
// (at your option) any later version.

// Astar is distributed in the hope that it will be useful,
// but WITHOUT ANY WARRANTY; without even the implied warranty of
// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
// GNU General Public License for more details.

// You should have received a copy of the GNU General Public License
// along with Astar. If not, see <http://www.gnu.org/licenses/>.

//! Parachain Service and ServiceFactory implementation.

use astar_primitives::*;
use cumulus_client_bootnodes::{start_bootnode_tasks, StartBootnodeTasksParams};
use cumulus_client_cli::CollatorOptions;
use cumulus_client_consensus_aura::collators::slot_based::{
    self as aura, Params as AuraParams, SlotBasedBlockImport, SlotBasedBlockImportHandle,
};
use cumulus_client_consensus_common::ParachainBlockImport as TParachainBlockImport;
use cumulus_client_consensus_relay_chain::Verifier as RelayChainVerifier;
use cumulus_client_service::ParachainTracingExecuteBlock;
use cumulus_client_service::{
    prepare_node_config, start_relay_chain_tasks, BuildNetworkParams, DARecoveryProfile,
    StartRelayChainTasksParams,
};
use cumulus_primitives_core::{
    relay_chain::{CollatorPair, ValidationCode},
    ParaId,
};
use cumulus_relay_chain_interface::RelayChainInterface;
use fc_consensus::FrontierBlockImport as TFrontierBlockImport;
use fc_rpc_core::types::{FeeHistoryCache, FilterPool};
use fc_storage::StorageOverrideHandler;
use futures::StreamExt;
use sc_client_api::BlockchainEvents;
use sc_client_db::PruningMode;
use sc_consensus::{import_queue::BasicQueue, ImportQueue};
use sc_executor::{HeapAllocStrategy, WasmExecutor, DEFAULT_HEAP_ALLOC_STRATEGY};
use sc_network::{config::NetworkBackendType, NetworkBackend, NetworkBlock, PeerId};
use sc_network_sync::SyncingService;
use sc_service::{Configuration, PartialComponents, TFullBackend, TFullClient, TaskManager};
use sc_telemetry::{Telemetry, TelemetryHandle, TelemetryWorker, TelemetryWorkerHandle};
use sp_api::{ApiExt, ProvideRuntimeApi};
use sp_consensus_aura::{
    sr25519::AuthorityId as AuraId, sr25519::AuthorityPair as AuraPair, AuraApi,
};
use sp_keystore::KeystorePtr;
use sp_runtime::{traits::Block as BlockT, Percent};
use std::{collections::BTreeMap, sync::Arc, time::Duration};
use substrate_prometheus_endpoint::Registry;

use super::shell_upgrade::*;

use crate::{
    evm_tracing_types::{EthApi as EthApiCmd, FrontierConfig},
    rpc::tracing,
};

/// Parachain host functions
#[cfg(feature = "runtime-benchmarks")]
pub type HostFunctions = (
    frame_benchmarking::benchmarking::HostFunctions,
    cumulus_client_service::ParachainHostFunctions,
    moonbeam_primitives_ext::moonbeam_ext::HostFunctions,
);

/// Parachain host functions
#[cfg(not(feature = "runtime-benchmarks"))]
pub type HostFunctions = (
    cumulus_client_service::ParachainHostFunctions,
    moonbeam_primitives_ext::moonbeam_ext::HostFunctions,
);

/// Parachain executor
pub type ParachainExecutor = WasmExecutor<HostFunctions>;

type FullClient =
    TFullClient<Block, crate::parachain::fake_runtime_api::RuntimeApi, ParachainExecutor>;

/// Frontier block import, wrapping the inner client.
type FrontierBlockImportType = TFrontierBlockImport<Block, Arc<FullClient>, FullClient>;

/// Slot-based block import sitting on top of Frontier.
type SlotBasedImport = SlotBasedBlockImport<Block, FrontierBlockImportType, FullClient>;

/// The full parachain block-import.
type ParachainBlockImport = TParachainBlockImport<Block, SlotBasedImport, TFullBackend<Block>>;

/// Starts a `ServiceBuilder` for a full service.
///
/// Use this macro if you don't actually need the full service, but just the builder in order to
/// be able to perform chain operations.
pub fn new_partial(
    config: &Configuration,
    evm_tracing_config: &FrontierConfig,
) -> Result<
    PartialComponents<
        FullClient,
        TFullBackend<Block>,
        (),
        sc_consensus::DefaultImportQueue<Block>,
        sc_transaction_pool::TransactionPoolHandle<Block, FullClient>,
        (
            ParachainBlockImport,
            SlotBasedBlockImportHandle<Block>,
            Option<Telemetry>,
            Option<TelemetryWorkerHandle>,
            Arc<fc_db::Backend<Block, FullClient>>,
        ),
    >,
    sc_service::Error,
> {
    let telemetry = config
        .telemetry_endpoints
        .clone()
        .filter(|x| !x.is_empty())
        .map(|endpoints| -> Result<_, sc_telemetry::Error> {
            let worker = TelemetryWorker::new(16)?;
            let telemetry = worker.handle().new_telemetry(endpoints);
            Ok((worker, telemetry))
        })
        .transpose()?;

    let heap_pages = config
        .executor
        .default_heap_pages
        .map_or(DEFAULT_HEAP_ALLOC_STRATEGY, |h| HeapAllocStrategy::Static {
            extra_pages: h as _,
        });

    let executor = ParachainExecutor::builder()
        .with_execution_method(config.executor.wasm_method)
        .with_onchain_heap_alloc_strategy(heap_pages)
        .with_offchain_heap_alloc_strategy(heap_pages)
        .with_max_runtime_instances(config.executor.max_runtime_instances)
        .with_runtime_cache_size(config.executor.runtime_cache_size)
        .build();

    let (client, backend, keystore_container, task_manager) =
        sc_service::new_full_parts_record_import::<Block, _, _>(
            config,
            telemetry.as_ref().map(|(_, telemetry)| telemetry.handle()),
            executor,
            true,
            Default::default(),
        )?;
    let client = Arc::new(client);

    let telemetry_worker_handle = telemetry.as_ref().map(|(worker, _)| worker.handle());

    let telemetry = telemetry.map(|(worker, telemetry)| {
        task_manager
            .spawn_handle()
            .spawn("telemetry", None, worker.run());
        telemetry
    });

    let transaction_pool = sc_transaction_pool::Builder::new(
        task_manager.spawn_essential_handle(),
        client.clone(),
        config.role.is_authority().into(),
    )
    .with_options(config.transaction_pool.clone())
    .with_prometheus(config.prometheus_registry())
    .build();

    let frontier_backend = Arc::new(crate::rpc::open_frontier_backend(
        client.clone(),
        config,
        evm_tracing_config,
    )?);

    let frontier_block_import = TFrontierBlockImport::new(client.clone(), client.clone());
    let (slot_based_block_import, slot_based_import_handle) =
        SlotBasedBlockImport::new(frontier_block_import, client.clone());
    let parachain_block_import: ParachainBlockImport =
        ParachainBlockImport::new(slot_based_block_import, backend.clone());

    let import_queue = build_import_queue(
        client.clone(),
        parachain_block_import.clone(),
        config,
        telemetry.as_ref().map(|telemetry| telemetry.handle()),
        &task_manager,
    );

    let params = PartialComponents {
        backend,
        client,
        import_queue,
        keystore_container,
        task_manager,
        transaction_pool: transaction_pool.into(),
        select_chain: (),
        other: (
            parachain_block_import,
            slot_based_import_handle,
            telemetry,
            telemetry_worker_handle,
            frontier_backend,
        ),
    };

    Ok(params)
}

#[derive(Clone)]
/// To add additional config to start_xyz_node functions
pub struct AdditionalConfig {
    /// EVM tracing configuration
    pub evm_tracing_config: FrontierConfig,

    /// Whether EVM RPC be enabled
    pub enable_evm_rpc: bool,

    /// Maxium allowed block size limit to propose
    pub proposer_block_size_limit: usize,

    /// Soft deadline limit used by `Proposer`
    pub proposer_soft_deadline_percent: u8,

    /// Hardware benchmarks score
    pub hwbench: Option<sc_sysinfo::HwBench>,
}

/// Start a node with the given parachain `Configuration` and relay chain `Configuration`.
///
/// This is the actual implementation that is abstract over the executor and the runtime api.
#[sc_tracing::logging::prefix_logs_with("Parachain")]
async fn start_node_impl<N>(
    parachain_config: Configuration,
    polkadot_config: Configuration,
    collator_options: CollatorOptions,
    para_id: ParaId,
    additional_config: AdditionalConfig,
) -> sc_service::error::Result<(TaskManager, Arc<FullClient>)>
where
    N: NetworkBackend<Block, <Block as BlockT>::Hash>,
{
    let parachain_config = prepare_node_config(parachain_config);

    let PartialComponents {
        client,
        backend,
        mut task_manager,
        keystore_container,
        select_chain: _,
        import_queue,
        transaction_pool,
        other:
            (
                parachain_block_import,
                block_import_handle,
                mut telemetry,
                telemetry_worker_handle,
                frontier_backend,
            ),
    } = new_partial(&parachain_config, &additional_config.evm_tracing_config)?;

    let prometheus_registry = parachain_config.prometheus_registry().cloned();
    let net_config = sc_network::config::FullNetworkConfiguration::<_, _, N>::new(
        &parachain_config.network,
        prometheus_registry.clone(),
    );

    let metrics = N::register_notification_metrics(
        parachain_config
            .prometheus_config
            .as_ref()
            .map(|cfg| &cfg.registry),
    );

    // Extract relay_chain_fork_id before polkadot_config is moved into the builder.
    let relay_chain_fork_id = polkadot_config
        .chain_spec
        .fork_id()
        .map(ToString::to_string);

    let (relay_chain_interface, collator_key, relay_chain_network, paranode_rx) =
        cumulus_client_service::build_relay_chain_interface(
            polkadot_config,
            &parachain_config,
            telemetry_worker_handle,
            &mut task_manager,
            collator_options.clone(),
            additional_config.hwbench.clone(),
        )
        .await
        .map_err(|e| sc_service::Error::Application(Box::new(e) as Box<_>))?;

    let is_authority = parachain_config.role.is_authority();
    let import_queue_service = import_queue.service();
    let (network, system_rpc_tx, tx_handler_controller, sync_service) =
        cumulus_client_service::build_network(BuildNetworkParams {
            parachain_config: &parachain_config,
            net_config,
            para_id,
            client: client.clone(),
            transaction_pool: transaction_pool.clone(),
            spawn_handle: task_manager.spawn_handle(),
            spawn_essential_handle: task_manager.spawn_essential_handle(),
            import_queue,
            relay_chain_interface: relay_chain_interface.clone(),
            sybil_resistance_level: cumulus_client_service::CollatorSybilResistance::Resistant,
            metrics,
        })
        .await?;

    let filter_pool: FilterPool = Arc::new(std::sync::Mutex::new(BTreeMap::new()));
    let fee_history_cache: FeeHistoryCache = Arc::new(std::sync::Mutex::new(BTreeMap::new()));
    let storage_override = Arc::new(StorageOverrideHandler::new(client.clone()));

    // Sinks for pubsub notifications.
    // Everytime a new subscription is created, a new mpsc channel is added to the sink pool.
    // The MappingSyncWorker sends through the channel on block import and the subscription emits a notification to the subscriber on receiving a message through this channel.
    // This way we avoid race conditions when using native substrate block import notification stream.
    let pubsub_notification_sinks: fc_mapping_sync::EthereumBlockNotificationSinks<
        fc_mapping_sync::EthereumBlockNotification<Block>,
    > = Default::default();
    let pubsub_notification_sinks = Arc::new(pubsub_notification_sinks);

    let ethapi_cmd = additional_config.evm_tracing_config.ethapi.clone();
    let tracing_requesters =
        if ethapi_cmd.contains(&EthApiCmd::Debug) || ethapi_cmd.contains(&EthApiCmd::Trace) {
            tracing::spawn_tracing_tasks(
                &additional_config.evm_tracing_config,
                prometheus_registry.clone(),
                tracing::SpawnTasksParams {
                    task_manager: &task_manager,
                    client: client.clone(),
                    substrate_backend: backend.clone(),
                    frontier_backend: frontier_backend.clone(),
                    storage_override: storage_override.clone(),
                },
            )
        } else {
            tracing::RpcRequesters {
                debug: None,
                trace: None,
            }
        };

    // Frontier offchain DB task. Essential.
    // Maps emulated ethereum data to substrate native data.
    match frontier_backend.as_ref() {
        fc_db::Backend::KeyValue(ref b) => {
            task_manager.spawn_essential_handle().spawn(
                "frontier-mapping-sync-worker",
                Some("frontier"),
                fc_mapping_sync::kv::MappingSyncWorker::new(
                    client.import_notification_stream(),
                    Duration::new(6, 0),
                    client.clone(),
                    backend.clone(),
                    storage_override.clone(),
                    b.clone(),
                    3,
                    0,
                    parachain_config.state_pruning.clone().and_then(|mode| {
                        if let PruningMode::Constrained(c) = mode {
                            c.max_blocks.map(u64::from)
                        } else {
                            None
                        }
                    }),
                    fc_mapping_sync::SyncStrategy::Parachain,
                    sync_service.clone(),
                    pubsub_notification_sinks.clone(),
                )
                .for_each(|()| futures::future::ready(())),
            );
        }
        fc_db::Backend::Sql(ref b) => {
            task_manager.spawn_essential_handle().spawn_blocking(
                "frontier-mapping-sync-worker",
                Some("frontier"),
                fc_mapping_sync::sql::SyncWorker::run(
                    client.clone(),
                    backend.clone(),
                    b.clone(),
                    client.import_notification_stream(),
                    fc_mapping_sync::sql::SyncWorkerConfig {
                        read_notification_timeout: Duration::from_secs(10),
                        check_indexed_blocks_interval: Duration::from_secs(60),
                    },
                    fc_mapping_sync::SyncStrategy::Parachain,
                    sync_service.clone(),
                    pubsub_notification_sinks.clone(),
                ),
            );
        }
    }

    // Frontier `EthFilterApi` maintenance. Manages the pool of user-created Filters.
    // Each filter is allowed to stay in the pool for 100 blocks.
    const FILTER_RETAIN_THRESHOLD: u64 = 100;
    task_manager.spawn_essential_handle().spawn(
        "frontier-filter-pool",
        Some("frontier"),
        fc_rpc::EthTask::filter_pool_task(
            client.clone(),
            filter_pool.clone(),
            FILTER_RETAIN_THRESHOLD,
        ),
    );

    const FEE_HISTORY_LIMIT: u64 = 2048;
    task_manager.spawn_essential_handle().spawn(
        "frontier-fee-history",
        Some("frontier"),
        fc_rpc::EthTask::fee_history_task(
            client.clone(),
            storage_override.clone(),
            fee_history_cache.clone(),
            FEE_HISTORY_LIMIT,
        ),
    );

    let block_data_cache = Arc::new(fc_rpc::EthBlockDataCacheTask::new(
        task_manager.spawn_handle(),
        storage_override.clone(),
        50,
        50,
        prometheus_registry.clone(),
    ));

    let rpc_extensions_builder = {
        let client = client.clone();
        let network = network.clone();
        let transaction_pool = transaction_pool.clone();
        let rpc_config = crate::rpc::EvmTracingConfig {
            tracing_requesters,
            trace_filter_max_count: additional_config.evm_tracing_config.ethapi_trace_max_count,
            trace_filter_max_block_range: additional_config
                .evm_tracing_config
                .trace_filter_max_block_range,
            enable_txpool: ethapi_cmd.contains(&EthApiCmd::TxPool),
        };
        let sync = sync_service.clone();
        let pubsub_notification_sinks = pubsub_notification_sinks.clone();

        Box::new(move |subscription| {
            let deps = crate::rpc::FullDeps {
                client: client.clone(),
                pool: transaction_pool.clone(),
                graph: transaction_pool.clone(),
                network: network.clone(),
                sync: sync.clone(),
                is_authority,
                frontier_backend: match *frontier_backend {
                    fc_db::Backend::KeyValue(ref b) => b.clone(),
                    fc_db::Backend::Sql(ref b) => b.clone(),
                },
                filter_pool: filter_pool.clone(),
                fee_history_limit: FEE_HISTORY_LIMIT,
                fee_history_cache: fee_history_cache.clone(),
                block_data_cache: block_data_cache.clone(),
                storage_override: storage_override.clone(),
                enable_evm_rpc: additional_config.enable_evm_rpc,
                command_sink: None,
            };

            crate::rpc::create_full(
                deps,
                subscription,
                pubsub_notification_sinks.clone(),
                rpc_config.clone(),
            )
            .map_err(Into::into)
        })
    };

    // Extract values needed after `parachain_config` is moved into `spawn_tasks`.
    let parachain_advertise_non_global_ips = parachain_config.network.allow_non_globals_in_dht;
    let parachain_fork_id = parachain_config
        .chain_spec
        .fork_id()
        .map(ToString::to_string);
    let parachain_public_addresses = parachain_config.network.public_addresses.clone();

    // Spawn basic services.
    sc_service::spawn_tasks(sc_service::SpawnTasksParams {
        rpc_builder: rpc_extensions_builder,
        client: client.clone(),
        transaction_pool: transaction_pool.clone(),
        task_manager: &mut task_manager,
        config: parachain_config,
        keystore: keystore_container.keystore(),
        backend: backend.clone(),
        network: network.clone(),
        system_rpc_tx,
        sync_service: sync_service.clone(),
        tx_handler_controller,
        telemetry: telemetry.as_mut(),
        tracing_execute_block: Some(Arc::new(ParachainTracingExecuteBlock::new(client.clone()))),
    })?;

    if let Some(hwbench) = additional_config.hwbench.clone() {
        sc_sysinfo::print_hwbench(&hwbench);
        if is_authority {
            warn_if_slow_hardware(&hwbench);
        }

        if let Some(ref mut telemetry) = telemetry {
            let telemetry_handle = telemetry.handle();
            task_manager.spawn_handle().spawn(
                "telemetry_hwbench",
                None,
                sc_sysinfo::initialize_hwbench_telemetry(telemetry_handle, hwbench),
            );
        }
    }

    let announce_block = {
        let sync_service = sync_service.clone();
        Arc::new(move |hash, data| sync_service.announce_block(hash, data))
    };

    let overseer_handle = relay_chain_interface
        .overseer_handle()
        .map_err(|e| sc_service::Error::Application(Box::new(e)))?;

    start_relay_chain_tasks(StartRelayChainTasksParams {
        client: client.clone(),
        announce_block: announce_block.clone(),
        task_manager: &mut task_manager,
        para_id,
        relay_chain_interface: relay_chain_interface.clone(),
        relay_chain_slot_duration: Duration::from_secs(6),
        import_queue: import_queue_service,
        recovery_handle: Box::new(overseer_handle.clone()),
        sync_service: sync_service.clone(),
        da_recovery_profile: if is_authority {
            DARecoveryProfile::Collator
        } else {
            DARecoveryProfile::FullNode
        },
        prometheus_registry: prometheus_registry.as_ref(),
    })?;

    start_bootnode_tasks(StartBootnodeTasksParams {
        embedded_dht_bootnode: collator_options.embedded_dht_bootnode,
        dht_bootnode_discovery: collator_options.dht_bootnode_discovery,
        para_id,
        task_manager: &mut task_manager,
        relay_chain_interface: relay_chain_interface.clone(),
        relay_chain_fork_id,
        relay_chain_network,
        request_receiver: paranode_rx,
        parachain_network: network.clone(),
        advertise_non_global_ips: parachain_advertise_non_global_ips,
        parachain_genesis_hash: client.chain_info().genesis_hash.as_ref().to_vec(),
        parachain_fork_id,
        parachain_public_addresses,
    });

    if is_authority {
        start_aura_consensus(
            client.clone(),
            backend,
            parachain_block_import,
            block_import_handle,
            prometheus_registry.as_ref(),
            telemetry.map(|t| t.handle()),
            &mut task_manager,
            relay_chain_interface,
            transaction_pool,
            sync_service,
            keystore_container.keystore(),
            para_id,
            collator_key.expect("Command line arguments do not allow this. qed"),
            network.local_peer_id(),
            additional_config,
        )?;
    }

    Ok((task_manager, client))
}

/// Build aura import queue with fallback to relay-chain verifier.
/// Starts with relay-chain verifier until aura becomes available.
pub fn build_import_queue(
    client: Arc<FullClient>,
    block_import: ParachainBlockImport,
    config: &Configuration,
    telemetry_handle: Option<TelemetryHandle>,
    task_manager: &TaskManager,
) -> sc_consensus::DefaultImportQueue<Block> {
    let verifier_client = client.clone();
    // CIDP for Aura verifier - requires AuraApi for slot_duration
    let create_aura_inherent_data_providers = move |parent_hash, _| {
        let cidp_client = verifier_client.clone();
        async move {
            let slot_duration =
                cumulus_client_consensus_aura::slot_duration_at(&*cidp_client, parent_hash)?;
            let timestamp = sp_timestamp::InherentDataProvider::from_system_time();

            let slot =
                sp_consensus_aura::inherents::InherentDataProvider::from_timestamp_and_slot_duration(
                    *timestamp,
                    slot_duration,
                );

            Ok((slot, timestamp))
        }
    };

    // CIDP for relay chain verifier - minimal, only timestamp (no AuraApi required)
    // This is used for pre-AuraApi blocks (Shiden genesis did not start with AuraApi)
    let create_relay_inherent_data_providers = move |_parent_hash: Hash, _| async move {
        let timestamp = sp_timestamp::InherentDataProvider::from_system_time();
        Ok(timestamp)
    };

    let aura_verifier = Box::new(cumulus_client_consensus_aura::build_verifier::<
        AuraPair,
        _,
        _,
        _,
    >(cumulus_client_consensus_aura::BuildVerifierParams {
        client: client.clone(),
        create_inherent_data_providers: create_aura_inherent_data_providers,
        telemetry: telemetry_handle,
    }));

    let relay_chain_verifier = Box::new(RelayChainVerifier::new(
        client.clone(),
        create_relay_inherent_data_providers,
    )) as Box<_>;

    let verifier = Verifier {
        client,
        relay_chain_verifier,
        aura_verifier,
    };

    let registry = config.prometheus_registry();
    let spawner = task_manager.spawn_essential_handle();

    BasicQueue::new(verifier, Box::new(block_import), None, &spawner, registry)
}

/// Start collating with the `shell` runtime while waiting for an upgrade to an Aura compatible runtime.
fn start_aura_consensus(
    client: Arc<FullClient>,
    backend: Arc<TFullBackend<Block>>,
    block_import: ParachainBlockImport,
    block_import_handle: SlotBasedBlockImportHandle<Block>,
    prometheus_registry: Option<&Registry>,
    telemetry: Option<TelemetryHandle>,
    task_manager: &TaskManager,
    relay_chain_interface: Arc<dyn RelayChainInterface>,
    transaction_pool: Arc<sc_transaction_pool::TransactionPoolHandle<Block, FullClient>>,
    sync_oracle: Arc<SyncingService<Block>>,
    keystore: KeystorePtr,
    para_id: ParaId,
    collator_key: CollatorPair,
    collator_peer_id: PeerId,
    additional_config: AdditionalConfig,
) -> Result<(), sc_service::Error> {
    let mut proposer_factory = sc_basic_authorship::ProposerFactory::new(
        task_manager.spawn_handle(),
        client.clone(),
        transaction_pool,
        prometheus_registry,
        telemetry,
    );

    proposer_factory.set_default_block_size_limit(additional_config.proposer_block_size_limit);
    proposer_factory.set_soft_deadline(Percent::from_percent(
        additional_config.proposer_soft_deadline_percent,
    ));

    let announce_block = {
        let sync_service = sync_oracle.clone();
        Arc::new(move |hash, data| sync_service.announce_block(hash, data))
    };

    let collator_service = cumulus_client_collator::service::CollatorService::new(
        client.clone(),
        Arc::new(task_manager.spawn_handle()),
        announce_block,
        client.clone(),
    );

    let params = AuraParams {
        create_inherent_data_providers: move |_, ()| async move { Ok(()) },
        block_import,
        para_client: client.clone(),
        para_backend: backend,
        relay_client: relay_chain_interface.clone(),
        code_hash_provider: {
            let client = client.clone();
            move |block_hash| {
                client
                    .code_at(block_hash)
                    .ok()
                    .map(|c| ValidationCode::from(c).hash())
            }
        },
        keystore,
        collator_key,
        collator_peer_id,
        para_id,
        slot_offset: Duration::from_secs(1),
        relay_chain_slot_duration: Duration::from_secs(6),
        proposer: proposer_factory,
        collator_service,
        authoring_duration: Duration::from_millis(2000),
        reinitialize: false,
        block_import_handle,
        spawner: task_manager.spawn_essential_handle(),
        // If necessary, AdditionalConfig CLI params could be extend to make it configurable.
        // However, it will be removed once https://github.com/paritytech/polkadot-sdk/issues/6020 is fixed.
        max_pov_percentage: None, // default is 85%
        export_pov: None,
    };

    let fut = async move {
        wait_for_aura(client).await;
        aura::run::<Block, AuraPair, _, _, _, _, _, _, _, _, _>(params);
    };

    task_manager.spawn_handle().spawn("aura", None, fut);
    Ok(())
}

/// Wait for the Aura runtime API to appear on chain.
/// This is useful for chains that started out without Aura. Components that
/// are depending on Aura functionality will wait until Aura appears in the runtime.
async fn wait_for_aura(client: Arc<FullClient>) {
    let finalized_hash = client.chain_info().finalized_hash;
    if client
        .runtime_api()
        .has_api::<dyn AuraApi<Block, AuraId>>(finalized_hash)
        .unwrap_or_default()
    {
        return;
    };

    let mut stream = client.finality_notification_stream();
    while let Some(notification) = stream.next().await {
        if client
            .runtime_api()
            .has_api::<dyn AuraApi<Block, AuraId>>(notification.hash)
            .unwrap_or_default()
        {
            return;
        }
    }
}

/// Checks that the hardware meets the requirements and print a warning otherwise.
fn warn_if_slow_hardware(hwbench: &sc_sysinfo::HwBench) {
    // Polkadot para-chains should generally use these requirements to ensure that the relay-chain
    // will not take longer than expected to import its blocks.
    if let Err(err) =
        frame_benchmarking_cli::SUBSTRATE_REFERENCE_HARDWARE.check_hardware(hwbench, false)
    {
        log::warn!(
            "⚠️  The hardware does not meet the minimal requirements {} for role 'Authority' find out more at:\n\
            https://wiki.polkadot.network/docs/maintain-guides-how-to-validate-polkadot#reference-hardware",
            err
        );
    }
}

/// Start a parachain node.
pub async fn start_node(
    parachain_config: Configuration,
    polkadot_config: Configuration,
    collator_options: CollatorOptions,
    para_id: ParaId,
    additional_config: AdditionalConfig,
) -> sc_service::error::Result<(TaskManager, Arc<FullClient>)> {
    match parachain_config.network.network_backend {
        NetworkBackendType::Libp2p => {
            start_node_impl::<sc_network::NetworkWorker<_, _>>(
                parachain_config,
                polkadot_config,
                collator_options,
                para_id,
                additional_config,
            )
            .await
        }
        NetworkBackendType::Litep2p => {
            start_node_impl::<sc_network::Litep2pNetworkBackend>(
                parachain_config,
                polkadot_config,
                collator_options,
                para_id,
                additional_config,
            )
            .await
        }
    }
}