smeb_y/0g-storage-node
1
0
Fork 0
mirror of https://github.com/0glabs/0g-storage-node.git synced 2024-12-26 00:05:17 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity

Repeat log catch-up until it's close to the latest height. (#114)

Browse Source 
* Repeat log catch-up until it's close to the latest height.

* Add an oneshot channel to send catch up end complete.

* Fix comments.

* Fix tests.

* Fix ut.
This commit is contained in:
peilun-conflux 2024-07-11 14:07:03 +08:00 committed by GitHub
parent 59d3a03487
commit 3e23e81ae6
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
4 changed files with 159 additions and 122 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

237
node/log_entry_sync/src/sync_manager/mod.rs
View File

@ -2,6 +2,7 @@ use crate::sync_manager::config::LogSyncConfig;
use crate::sync_manager::data_cache::DataCache;
use crate::sync_manager::log_entry_fetcher::{LogEntryFetcher, LogFetchProgress};
use anyhow::{anyhow, bail, Result};
use ethereum_types::H256;
use ethers::{prelude::Middleware, types::BlockNumber};
use futures::FutureExt;
use jsonrpsee::tracing::{debug, error, trace, warn};
@ -15,10 +16,11 @@ use storage::log_store::{tx_store::BlockHashAndSubmissionIndex, Store};
use task_executor::{ShutdownReason, TaskExecutor};
use tokio::sync::broadcast;
use tokio::sync::mpsc::UnboundedReceiver;
use tokio::sync::RwLock;
use tokio::sync::{oneshot, RwLock};
const RETRY_WAIT_MS: u64 = 500;
const BROADCAST_CHANNEL_CAPACITY: usize = 16;
const CATCH_UP_END_GAP: u64 = 10;
#[derive(Clone, Debug)]
pub enum LogSyncEvent {
@ -49,7 +51,7 @@ impl LogSyncManager {
config: LogSyncConfig,
executor: TaskExecutor,
store: Arc<dyn Store>,
) -> Result<broadcast::Sender<LogSyncEvent>> {
) -> Result<(broadcast::Sender<LogSyncEvent>, oneshot::Receiver<()>)> {
let next_tx_seq = store.next_tx_seq();
let executor_clone = executor.clone();
@ -57,6 +59,7 @@ impl LogSyncManager {
let (event_send, _) = broadcast::channel(BROADCAST_CHANNEL_CAPACITY);
let event_send_cloned = event_send.clone();
let (catch_up_end_sender, catch_up_end_receiver) = oneshot::channel();
// Spawn the task to sync log entries from the blockchain.
executor.spawn(
@ -96,69 +99,38 @@ impl LogSyncManager {
block_hash_cache,
};
let finalized_block = match log_sync_manager
.log_fetcher
.provider()
.get_block(BlockNumber::Finalized)
.await
{
Ok(Some(finalized_block)) => finalized_block,
e => {
warn!("unable to get finalized block: {:?}", e);
log_sync_manager
.log_fetcher
.provider()
.get_block(0)
.await?
.ok_or_else(|| anyhow!("None for block 0"))?
}
};
let finalized_block_number = finalized_block
.number
.ok_or_else(|| anyhow!("None block number for finalized block"))?
.as_u64();
// Load previous progress from db and check if chain reorg happens after restart.
// TODO(zz): Handle reorg instead of return.
let mut need_handle_reorg = false;
let (mut start_block_number, mut start_block_hash) =
match log_sync_manager.store.get_sync_progress()? {
// No previous progress, so just use config.
None => {
let block_number = log_sync_manager.config.start_block_number;
let block_hash = log_sync_manager
.log_fetcher
.provider()
.get_block(block_number)
.await?
.ok_or_else(|| anyhow!("None for block {}", block_number))?
.hash
.ok_or_else(|| {
anyhow!("None block hash for block {}", block_number)
})?;
let block_hash =
log_sync_manager.get_block(block_number.into()).await?.1;
(block_number, block_hash)
}
Some((block_number, block_hash)) => {
if block_number <= finalized_block_number {
let expect_block_hash = log_sync_manager
.log_fetcher
.provider()
.get_block(block_number)
.await?
.ok_or_else(|| anyhow!("None for block {}", block_number))?
.hash
.ok_or_else(|| {
anyhow!("None block hash for block {}", block_number)
})?;
Some((block_number, block_hash)) => (block_number, block_hash),
};
if expect_block_hash != block_hash {
need_handle_reorg = true;
}
}
(block_number, block_hash)
let (mut finalized_block_number, mut finalized_block_hash) =
match log_sync_manager.get_block(BlockNumber::Finalized).await {
Ok(finalized) => finalized,
Err(e) => {
warn!(?e, "unable to get finalized block");
log_sync_manager.get_block(0.into()).await?
}
};
// Load previous progress from db and check if chain reorg happens after restart.
let mut need_handle_reorg = false;
if start_block_number <= finalized_block_number {
let expect_block_hash = log_sync_manager
.get_block(start_block_number.into())
.await?
.1;
if expect_block_hash != start_block_hash {
need_handle_reorg = true;
}
}
debug!(
"current start block number {}, block hash {}, finalized block number {}",
start_block_number, start_block_hash, finalized_block_number
@ -172,7 +144,6 @@ impl LogSyncManager {
log_sync_manager.block_hash_cache.clone(),
);
log_sync_manager.handle_data(reorg_rx).await?;
if let Some((block_number, block_hash)) =
log_sync_manager.store.get_sync_progress()?
{
@ -183,21 +154,26 @@ impl LogSyncManager {
}
}
// Start watching before recovery to ensure that no log is skipped.
// TODO(zz): Rate limit to avoid OOM during recovery.
let mut submission_idx = None;
let parent_block_hash = if start_block_number >= finalized_block_number {
if start_block_number > 0 {
if let Some(b) = log_sync_manager
.block_hash_cache
.read()
.await
.get(&start_block_number)
if start_block_number >= finalized_block_number {
let block = log_sync_manager
.block_hash_cache
.read()
.await
.get(&start_block_number)
.cloned();
if let Some(b) = block {
// special case avoid reorg
if let Some(submission_idx) = b.as_ref().unwrap().first_submission_index
{
// special case avoid reorg
submission_idx = b.as_ref().unwrap().first_submission_index;
log_sync_manager.process_reverted(submission_idx).await;
}
}
}
let parent_block_hash = if start_block_number >= finalized_block_number {
// No need to catch up data.
if start_block_number > 0 {
let parent_block_number = start_block_number.saturating_sub(1);
match log_sync_manager
.block_hash_cache
@ -206,75 +182,60 @@ impl LogSyncManager {
.get(&parent_block_number)
{
Some(b) => b.as_ref().unwrap().block_hash,
_ => log_sync_manager
.log_fetcher
.provider()
.get_block(parent_block_number)
.await?
.ok_or_else(|| {
anyhow!("None for block {}", parent_block_number)
})?
.hash
.ok_or_else(|| {
anyhow!("None block hash for block {}", parent_block_number)
})?,
_ => {
log_sync_manager
.get_block(parent_block_number.into())
.await?
.1
}
}
} else {
start_block_hash
}
} else {
finalized_block
.hash
.ok_or_else(|| anyhow!("None for finalized block hash"))?
// Keep catching-up data until we are close to the latest height.
loop {
log_sync_manager
.catch_up_data(
executor_clone.clone(),
start_block_number,
finalized_block_number,
)
.await?;
start_block_number = finalized_block_number.saturating_add(1);
let new_finalized_block =
log_sync_manager.get_block(BlockNumber::Finalized).await?;
if new_finalized_block.0.saturating_sub(finalized_block_number)
<= CATCH_UP_END_GAP
{
break finalized_block_hash;
}
finalized_block_number = new_finalized_block.0;
finalized_block_hash = new_finalized_block.1;
}
};
if let Some(submission_idx) = submission_idx {
log_sync_manager.process_reverted(submission_idx).await;
if catch_up_end_sender.send(()).is_err() {
warn!("catch_up_end send fails, possibly auto_sync is not enabled");
}
let watch_rx = log_sync_manager.log_fetcher.start_watch(
if start_block_number >= finalized_block_number {
start_block_number
} else {
finalized_block_number.saturating_add(1)
},
start_block_number,
parent_block_hash,
&executor_clone,
log_sync_manager.block_hash_cache.clone(),
log_sync_manager.config.watch_loop_wait_time_ms,
);
if start_block_number < finalized_block_number {
let recover_rx = log_sync_manager.log_fetcher.start_recover(
start_block_number,
finalized_block_number,
&executor_clone,
Duration::from_millis(log_sync_manager.config.recover_query_delay),
);
log_sync_manager.handle_data(recover_rx).await?;
}
log_sync_manager
.log_fetcher
.start_remove_finalized_block_task(
&executor_clone,
log_sync_manager.store.clone(),
log_sync_manager.block_hash_cache.clone(),
log_sync_manager.config.default_finalized_block_count,
log_sync_manager
.config
.remove_finalized_block_interval_minutes,
);
// Syncing `watch_rx` is supposed to block forever.
log_sync_manager.handle_data(watch_rx).await?;
Ok(())
Ok::<(), anyhow::Error>(())
},
)
.map(|_| ()),
"log_sync",
);
Ok(event_send_cloned)
Ok((event_send_cloned, catch_up_end_receiver))
}
async fn put_tx(&mut self, tx: Transaction) -> bool {
@ -423,6 +384,54 @@ impl LogSyncManager {
true
}
}
async fn get_block(&self, block_number: BlockNumber) -> Result<(u64, H256)> {
let block = match self.log_fetcher.provider().get_block(block_number).await {
Ok(Some(block)) => block,
Ok(None) => {
bail!("None for block {}", block_number);
}
e => {
bail!("unable to get block: {:?}", e);
}
};
Ok((
block
.number
.ok_or_else(|| anyhow!("None block number for finalized block"))?
.as_u64(),
block
.hash
.ok_or_else(|| anyhow!("None block hash for finalized block"))?,
))
}
/// Return the ending block number and the parent block hash.
async fn catch_up_data(
&mut self,
executor_clone: TaskExecutor,
start_block_number: u64,
finalized_block_number: u64,
) -> Result<()> {
if start_block_number < finalized_block_number {
let recover_rx = self.log_fetcher.start_recover(
start_block_number,
finalized_block_number,
&executor_clone,
Duration::from_millis(self.config.recover_query_delay),
);
self.handle_data(recover_rx).await?;
}
self.log_fetcher.start_remove_finalized_block_task(
&executor_clone,
self.store.clone(),
self.block_hash_cache.clone(),
self.config.default_finalized_block_count,
self.config.remove_finalized_block_interval_minutes,
);
Ok(())
}
}
async fn run_and_log<R, E>(

18
node/src/client/builder.rs
View File

@ -15,7 +15,7 @@ use storage::log_store::log_manager::LogConfig;
use storage::log_store::Store;
use storage::{LogManager, StorageConfig};
use sync::{SyncSender, SyncService};
use tokio::sync::{broadcast, mpsc};
use tokio::sync::{broadcast, mpsc, oneshot};
macro_rules! require {
($component:expr, $self:ident, $e:ident) => {
@ -48,6 +48,7 @@ struct MinerComponents {
struct LogSyncComponents {
send: broadcast::Sender<LogSyncEvent>,
catch_up_end_recv: Option<oneshot::Receiver<()>>,
}
struct PrunerComponents {
@ -163,6 +164,13 @@ impl ClientBuilder {
let file_location_cache = require!("sync", self, file_location_cache).clone();
let network_send = require!("sync", self, network).send.clone();
let event_recv = require!("sync", self, log_sync).send.subscribe();
let catch_up_end_recv = self
.log_sync
.as_mut()
.expect("checked in event_recv")
.catch_up_end_recv
.take()
.ok_or("sync requires a catch_up_end_recv")?;
let send = SyncService::spawn_with_config(
config,
@ -171,6 +179,7 @@ impl ClientBuilder {
store,
file_location_cache,
event_recv,
catch_up_end_recv,
)
.await
.map_err(|e| format!("Failed to start sync service: {:?}", e))?;
@ -295,11 +304,14 @@ impl ClientBuilder {
pub async fn with_log_sync(mut self, config: LogSyncConfig) -> Result<Self, String> {
let executor = require!("log_sync", self, runtime_context).clone().executor;
let store = require!("log_sync", self, store).clone();
let send = LogSyncManager::spawn(config, executor, store)
let (send, catch_up_end_recv) = LogSyncManager::spawn(config, executor, store)
.await
.map_err(|e| e.to_string())?;
self.log_sync = Some(LogSyncComponents { send });
self.log_sync = Some(LogSyncComponents {
send,
catch_up_end_recv: Some(catch_up_end_recv),
});
Ok(self)
}

4
node/sync/src/auto_sync/batcher_serial.rs
View File

@ -10,6 +10,7 @@ use anyhow::Result;
use log_entry_sync::LogSyncEvent;
use std::{collections::HashMap, fmt::Debug, sync::Arc};
use storage_async::Store;
use tokio::sync::oneshot;
use tokio::{
sync::{broadcast::Receiver, mpsc::UnboundedReceiver},
time::sleep,
@ -83,9 +84,12 @@ impl SerialBatcher {
mut self,
mut file_announcement_recv: UnboundedReceiver<u64>,
mut log_sync_recv: Receiver<LogSyncEvent>,
catch_up_end_recv: oneshot::Receiver<()>,
) {
info!(?self, "Start to sync files");
catch_up_end_recv.await.expect("log sync sender dropped");
loop {
// handle all pending file announcements
if self

22
node/sync/src/service.rs
View File

@ -26,7 +26,7 @@ use storage::error::Result as StorageResult;
use storage::log_store::Store as LogStore;
use storage_async::Store;
use tokio::sync::mpsc::{unbounded_channel, UnboundedSender};
use tokio::sync::{broadcast, mpsc};
use tokio::sync::{broadcast, mpsc, oneshot};
const HEARTBEAT_INTERVAL_SEC: u64 = 5;
@ -135,6 +135,7 @@ impl SyncService {
store: Arc<dyn LogStore>,
file_location_cache: Arc<FileLocationCache>,
event_recv: broadcast::Receiver<LogSyncEvent>,
catch_up_end_recv: oneshot::Receiver<()>,
) -> Result<SyncSender> {
Self::spawn_with_config(
Config::default(),
@ -143,6 +144,7 @@ impl SyncService {
store,
file_location_cache,
event_recv,
catch_up_end_recv,
)
.await
}
@ -154,6 +156,7 @@ impl SyncService {
store: Arc<dyn LogStore>,
file_location_cache: Arc<FileLocationCache>,
event_recv: broadcast::Receiver<LogSyncEvent>,
catch_up_end_recv: oneshot::Receiver<()>,
) -> Result<SyncSender> {
let (sync_send, sync_recv) = channel::Channel::unbounded();
@ -171,7 +174,10 @@ impl SyncService {
let serial_batcher =
SerialBatcher::new(config, store.clone(), sync_send.clone(), sync_store.clone())
.await?;
executor.spawn(serial_batcher.start(recv, event_recv), "auto_sync_serial");
executor.spawn(
serial_batcher.start(recv, event_recv, catch_up_end_recv),
"auto_sync_serial",
);
// sync randomly
let random_batcher =
@ -772,6 +778,7 @@ mod tests {
network_send: UnboundedSender<NetworkMessage>,
network_recv: UnboundedReceiver<NetworkMessage>,
event_send: broadcast::Sender<LogSyncEvent>,
catch_up_end_recv: Option<oneshot::Receiver<()>>,
}
impl Default for TestSyncRuntime {
@ -788,6 +795,7 @@ mod tests {
let init_peer_id = identity::Keypair::generate_ed25519().public().to_peer_id();
let (network_send, network_recv) = mpsc::unbounded_channel::<NetworkMessage>();
let (event_send, _) = broadcast::channel(16);
let (_, catch_up_end_recv) = oneshot::channel();
let tx_ids = txs.iter().take(seq_size).map(|tx| tx.id()).collect();
@ -803,16 +811,17 @@ mod tests {
network_send,
network_recv,
event_send,
catch_up_end_recv: Some(catch_up_end_recv),
}
}
async fn spawn_sync_service(&self, with_peer_store: bool) -> SyncSender {
async fn spawn_sync_service(&mut self, with_peer_store: bool) -> SyncSender {
self.spawn_sync_service_with_config(with_peer_store, Config::default())
.await
}
async fn spawn_sync_service_with_config(
&self,
&mut self,
with_peer_store: bool,
config: Config,
) -> SyncSender {
@ -829,6 +838,7 @@ mod tests {
store,
self.file_location_cache.clone(),
self.event_send.subscribe(),
self.catch_up_end_recv.take().unwrap(),
)
.await
.unwrap()
@ -1198,6 +1208,7 @@ mod tests {
let (network_send, mut network_recv) = mpsc::unbounded_channel::<NetworkMessage>();
let (_event_send, event_recv) = broadcast::channel(16);
let (_, catch_up_end_recv) = oneshot::channel();
let sync_send = SyncService::spawn_with_config(
Config::default(),
runtime.task_executor.clone(),
@ -1205,6 +1216,7 @@ mod tests {
store.clone(),
file_location_cache,
event_recv,
catch_up_end_recv,
)
.await
.unwrap();
@ -1544,7 +1556,7 @@ mod tests {
#[tokio::test]
async fn test_sync_status_unknown() {
let runtime = TestSyncRuntime::default();
let mut runtime = TestSyncRuntime::default();
let sync_send = runtime.spawn_sync_service(false).await;
assert!(matches!(