feat(torii-core): parallelization

bin/torii/src/main.rs

@@ -125,6 +125,10 @@
     /// Polling interval in ms
     #[arg(long, default_value = "500")]
     polling_interval: u64,
+
+    /// Max concurrent tasks
+    #[arg(long, default_value = "100")]
+    max_concurrent_tasks: usize,
 }
 
 #[tokio::main]
@@ -170,19 +174,19 @@
     let provider: Arc<_> = JsonRpcClient::new(HttpTransport::new(args.rpc)).into();
 
     // Get world address
-    let world = WorldContractReader::new(args.world_address, &provider);
+    let world = WorldContractReader::new(args.world_address, provider.clone());
 
     let db = Sql::new(pool.clone(), args.world_address).await?;
 
     let processors = Processors {
         event: generate_event_processors_map(vec![
-            Box::new(RegisterModelProcessor),
-            Box::new(StoreSetRecordProcessor),
-            Box::new(MetadataUpdateProcessor),
-            Box::new(StoreDelRecordProcessor),
-            Box::new(EventMessageProcessor),
-            Box::new(StoreUpdateRecordProcessor),
-            Box::new(StoreUpdateMemberProcessor),
+            Arc::new(RegisterModelProcessor),
+            Arc::new(StoreSetRecordProcessor),
+            Arc::new(MetadataUpdateProcessor),
+            Arc::new(StoreDelRecordProcessor),
+            Arc::new(EventMessageProcessor),
+            Arc::new(StoreUpdateRecordProcessor),
+            Arc::new(StoreUpdateMemberProcessor),
         ])?,
         transaction: vec![Box::new(StoreTransactionProcessor)],
         ..Processors::default()
@@ -193,9 +197,10 @@
     let mut engine = Engine::new(
         world,
         db.clone(),
-        &provider,
+        provider.clone(),
         processors,
         EngineConfig {
+            max_concurrent_tasks: args.max_concurrent_tasks,
             start_block: args.start_block,
             events_chunk_size: args.events_chunk_size,
             index_pending: args.index_pending,

crates/torii/core/src/cache.rs

@@ -113,6 +113,11 @@ impl ModelCache {
         Ok(model)
     }
 
+    pub async fn set(&self, selector: Felt, model: Model) {
+        let mut cache = self.cache.write().await;
+        cache.insert(selector, model);
+    }
+
     pub async fn clear(&self) {
         self.cache.write().await.clear();
     }

crates/torii/core/src/engine.rs

@@ -1,5 +1,6 @@
 use std::collections::{BTreeMap, HashMap};
 use std::fmt::Debug;
+use std::sync::Arc;
 use std::time::Duration;
 
 use anyhow::Result;
@@ -11,8 +12,11 @@
     TransactionReceiptWithBlockInfo, TransactionWithReceipt,
 };
 use starknet::providers::Provider;
+use starknet_crypto::poseidon_hash_many;
 use tokio::sync::broadcast::Sender;
 use tokio::sync::mpsc::Sender as BoundedSender;
+use tokio::sync::Semaphore;
+use tokio::task::JoinSet;
 use tokio::time::sleep;
 use tracing::{debug, error, info, trace, warn};
 
@@ -21,14 +25,14 @@
 use crate::sql::Sql;
 
 #[allow(missing_debug_implementations)]
-pub struct Processors<P: Provider + Send + Sync + std::fmt::Debug> {
+pub struct Processors<P: Provider + Send + Sync + std::fmt::Debug + 'static> {
     pub block: Vec<Box<dyn BlockProcessor<P>>>,
     pub transaction: Vec<Box<dyn TransactionProcessor<P>>>,
-    pub event: HashMap<Felt, Box<dyn EventProcessor<P>>>,
+    pub event: HashMap<Felt, Arc<dyn EventProcessor<P>>>,
     pub catch_all_event: Box<dyn EventProcessor<P>>,
 }
 
-impl<P: Provider + Send + Sync + std::fmt::Debug> Default for Processors<P> {
+impl<P: Provider + Send + Sync + std::fmt::Debug + 'static> Default for Processors<P> {
     fn default() -> Self {
         Self {
             block: vec![],
@@ -48,6 +52,7 @@
     pub start_block: u64,
     pub events_chunk_size: u64,
     pub index_pending: bool,
+    pub max_concurrent_tasks: usize,
 }
 
 impl Default for EngineConfig {
@@ -57,6 +62,7 @@
             start_block: 0,
             events_chunk_size: 1024,
             index_pending: true,
+            max_concurrent_tasks: 100,
         }
     }
 }
@@ -83,23 +89,32 @@
     pub block_number: u64,
 }
 
+#[derive(Debug)]
+pub struct ParallelizedEvent {
+    pub block_number: u64,
+    pub block_timestamp: u64,
+    pub event_id: String,
+    pub event: Event,
+}
+
 #[allow(missing_debug_implementations)]
-pub struct Engine<P: Provider + Send + Sync + std::fmt::Debug> {
-    world: WorldContractReader<P>,
+pub struct Engine<P: Provider + Send + Sync + std::fmt::Debug + 'static> {
+    world: Arc<WorldContractReader<P>>,
     db: Sql,
     provider: Box<P>,
-    processors: Processors<P>,
+    processors: Arc<Processors<P>>,
     config: EngineConfig,
     shutdown_tx: Sender<()>,
     block_tx: Option<BoundedSender<u64>>,
+    tasks: HashMap<Felt, Vec<ParallelizedEvent>>,
 }
 
 struct UnprocessedEvent {
     keys: Vec<String>,
     data: Vec<String>,
 }
 
-impl<P: Provider + Send + Sync + std::fmt::Debug> Engine<P> {
+impl<P: Provider + Send + Sync + std::fmt::Debug + 'static> Engine<P> {
     pub fn new(
         world: WorldContractReader<P>,
         db: Sql,
@@ -109,7 +124,16 @@
         shutdown_tx: Sender<()>,
         block_tx: Option<BoundedSender<u64>>,
     ) -> Self {
-        Self { world, db, provider: Box::new(provider), processors, config, shutdown_tx, block_tx }
+        Self {
+            world: Arc::new(world),
+            db,
+            provider: Box::new(provider),
+            processors: Arc::new(processors),
+            config,
+            shutdown_tx,
+            block_tx,
+            tasks: HashMap::new(),
+        }
     }
 
     pub async fn start(&mut self) -> Result<()> {
@@ -397,11 +421,14 @@
             }
         }
 
+        // Process parallelized events
+        self.process_tasks().await?;
+
         // Set the head to the last processed pending transaction
         // Head block number should still be latest block number
         self.db.set_head(data.block_number - 1, last_pending_block_world_tx, last_pending_block_tx);
-
         self.db.execute().await?;
+
         Ok(())
     }
 
@@ -436,18 +463,55 @@
             }
         }
 
-        // We return None for the pending_block_tx because our process_range
-        // gets only specific events from the world. so some transactions
-        // might get ignored and wont update the cursor.
-        // so once the sync range is done, we assume all of the tx of the block
-        // have been processed.
+        // Process parallelized events
+        self.process_tasks().await?;
 
         self.db.set_head(data.latest_block_number, None, None);
         self.db.execute().await?;
 
         Ok(())
     }
 
+    async fn process_tasks(&mut self) -> Result<()> {
+        // We use a semaphore to limit the number of concurrent tasks
+        let semaphore = Arc::new(Semaphore::new(self.config.max_concurrent_tasks));
+
+        // Run all tasks concurrently
+        let mut set = JoinSet::new();
+        for (task_id, events) in self.tasks.drain() {
+            let db = self.db.clone();
+            let world = self.world.clone();
+            let processors = self.processors.clone();
+            let semaphore = semaphore.clone();
+
+            set.spawn(async move {
+                let _permit = semaphore.acquire().await.unwrap();
+                let mut local_db = db.clone();
+                for ParallelizedEvent { event_id, event, block_number, block_timestamp } in events {
+                    if let Some(processor) = processors.event.get(&event.keys[0]) {
+                        debug!(target: LOG_TARGET, event_name = processor.event_key(), task_id = %task_id, "Processing parallelized event.");
+
+                        if let Err(e) = processor
+                            .process(&world, &mut local_db, block_number, block_timestamp, &event_id, &event)
+                            .await
+                        {
+                            error!(target: LOG_TARGET, event_name = processor.event_key(), error = %e, task_id = %task_id, "Processing parallelized event.");
+                        }
+                    }
+                }
+                Ok::<_, anyhow::Error>(local_db)
+            });
+        }
+
+        // Join all tasks
+        while let Some(result) = set.join_next().await {
+            let local_db = result??;
+            self.db.merge(local_db)?;
+        }
+
+        Ok(())
+    }
+
     async fn get_block_timestamp(&self, block_number: u64) -> Result<u64> {
         match self.provider.get_block_with_tx_hashes(BlockId::Number(block_number)).await? {
             MaybePendingBlockWithTxHashes::Block(block) => Ok(block.timestamp),
@@ -477,7 +541,7 @@
                 block_timestamp,
                 &event_id,
                 &event,
-                transaction_hash,
+                // transaction_hash,
             )
             .await?;
         }
@@ -527,7 +591,7 @@
                     block_timestamp,
                     &event_id,
                     event,
-                    *transaction_hash,
+                    // *transaction_hash,
                 )
                 .await?;
             }
@@ -587,9 +651,9 @@
         block_timestamp: u64,
         event_id: &str,
         event: &Event,
-        transaction_hash: Felt,
+        // transaction_hash: Felt,
     ) -> Result<()> {
-        self.db.store_event(event_id, event, transaction_hash, block_timestamp);
+        // self.db.store_event(event_id, event, transaction_hash, block_timestamp);
         let event_key = event.keys[0];
 
         let Some(processor) = self.processors.event.get(&event_key) else {
@@ -627,14 +691,30 @@
             return Ok(());
         };
 
-        // if processor.validate(event) {
-        if let Err(e) = processor
-            .process(&self.world, &mut self.db, block_number, block_timestamp, event_id, event)
-            .await
-        {
-            error!(target: LOG_TARGET, event_name = processor.event_key(), error = %e, "Processing event.");
+        let task_identifier = match processor.event_key().as_str() {
+            "StoreSetRecord" | "StoreUpdateRecord" | "StoreUpdateMember" | "StoreDelRecord" => {
+                poseidon_hash_many(&[event.data[0], event.data[1]])
+            }
+            _ => Felt::ZERO,
+        };
+
+        // if we have a task identifier, we queue the event to be parallelized
+        if task_identifier != Felt::ZERO {
+            self.tasks.entry(task_identifier).or_default().push(ParallelizedEvent {
+                event_id: event_id.to_string(),
+                event: event.clone(),
+                block_number,
+                block_timestamp,
+            });
+        } else {
+            // if we dont have a task identifier, we process the event immediately
+            if let Err(e) = processor
+                .process(&self.world, &mut self.db, block_number, block_timestamp, event_id, event)
+                .await
+            {
+                error!(target: LOG_TARGET, event_name = processor.event_key(), error = %e, "Processing event.");
+            }
         }
-        // }
 
         Ok(())
     }

crates/torii/core/src/processors/mod.rs

@@ -1,4 +1,5 @@
 use std::collections::HashMap;
+use std::sync::Arc;
 
 use anyhow::{Error, Result};
 use async_trait::async_trait;
@@ -23,7 +24,7 @@ const ENTITY_ID_INDEX: usize = 1;
 const NUM_KEYS_INDEX: usize = 2;
 
 #[async_trait]
-pub trait EventProcessor<P>
+pub trait EventProcessor<P>: Send + Sync
 where
     P: Provider + Sync,
 {
@@ -48,7 +49,7 @@ where
 }
 
 #[async_trait]
-pub trait BlockProcessor<P: Provider + Sync> {
+pub trait BlockProcessor<P: Provider + Sync>: Send + Sync {
     fn get_block_number(&self) -> String;
     async fn process(
         &self,
@@ -60,7 +61,7 @@ pub trait BlockProcessor<P: Provider + Sync> {
 }
 
 #[async_trait]
-pub trait TransactionProcessor<P: Provider + Sync> {
+pub trait TransactionProcessor<P: Provider + Sync>: Send + Sync {
     #[allow(clippy::too_many_arguments)]
     async fn process(
         &self,
@@ -75,8 +76,8 @@ pub trait TransactionProcessor<P: Provider + Sync> {
 
 /// Given a list of event processors, generate a map of event keys to the event processor
 pub fn generate_event_processors_map<P: Provider + Sync + Send>(
-    event_processor: Vec<Box<dyn EventProcessor<P>>>,
-) -> Result<HashMap<Felt, Box<dyn EventProcessor<P>>>> {
+    event_processor: Vec<Arc<dyn EventProcessor<P>>>,
+) -> Result<HashMap<Felt, Arc<dyn EventProcessor<P>>>> {
     let mut event_processors = HashMap::new();
 
     for processor in event_processor {