[Merged by Bors] - use state cache to optimise historical state lookup

beacon_node/src/cli.rs

@@ -511,6 +511,13 @@ pub fn cli_app<'a, 'b>() -> App<'a, 'b> {
                 .help("Specifies how many blocks the database should cache in memory [default: 5]")
                 .takes_value(true)
         )
+        .arg(
+            Arg::with_name("historic-state-cache-size")
+                .long("historic-state-cache-size")
+                .value_name("SIZE")
+                .help("Specifies how many states from the freezer database should cache in memory [default: 1]")
+                .takes_value(true)
+        )
         /*
          * Execution Layer Integration
          */

beacon_node/src/config.rs

@@ -380,6 +380,12 @@ pub fn get_config<E: EthSpec>(
             .map_err(|_| "block-cache-size is not a valid integer".to_string())?;
     }
 
+    if let Some(historic_state_cache_size) = cli_args.value_of("historic-state-cache-size") {
+        client_config.store.historic_state_cache_size = historic_state_cache_size
+            .parse()
+            .map_err(|_| "historic-state-cache-size is not a valid integer".to_string())?;
+    }
+
     client_config.store.compact_on_init = cli_args.is_present("compact-db");
     if let Some(compact_on_prune) = cli_args.value_of("auto-compact-db") {
         client_config.store.compact_on_prune = compact_on_prune

beacon_node/store/src/config.rs

@@ -7,6 +7,7 @@ use types::{EthSpec, MinimalEthSpec};
 pub const PREV_DEFAULT_SLOTS_PER_RESTORE_POINT: u64 = 2048;
 pub const DEFAULT_SLOTS_PER_RESTORE_POINT: u64 = 8192;
 pub const DEFAULT_BLOCK_CACHE_SIZE: usize = 5;
+pub const DEFAULT_HISTORIC_STATE_CACHE_SIZE: usize = 1;
 
 /// Database configuration parameters.
 #[derive(Debug, Clone, PartialEq, Eq, Serialize, Deserialize)]
@@ -17,6 +18,8 @@ pub struct StoreConfig {
     pub slots_per_restore_point_set_explicitly: bool,
     /// Maximum number of blocks to store in the in-memory block cache.
     pub block_cache_size: usize,
+    /// Maximum number of states from freezer database to store in the in-memory state cache.
+    pub historic_state_cache_size: usize,
     /// Whether to compact the database on initialization.
     pub compact_on_init: bool,
     /// Whether to compact the database during database pruning.
@@ -43,6 +46,7 @@ impl Default for StoreConfig {
             slots_per_restore_point: MinimalEthSpec::slots_per_historical_root() as u64,
             slots_per_restore_point_set_explicitly: false,
             block_cache_size: DEFAULT_BLOCK_CACHE_SIZE,
+            historic_state_cache_size: DEFAULT_HISTORIC_STATE_CACHE_SIZE,
             compact_on_init: false,
             compact_on_prune: true,
             prune_payloads: true,

beacon_node/store/src/hot_cold_store.rs

@@ -40,7 +40,7 @@ use std::sync::Arc;
 use std::time::Duration;
 use types::*;
 
-/// On-disk database that stores finalized states efficiently.
+/// On-disk database that stores fnalized states efficiently.
 ///
 /// Stores vector fields like the `block_roots` and `state_roots` separately, and only stores
 /// intermittent "restore point" states pre-finalization.
@@ -62,6 +62,8 @@ pub struct HotColdDB<E: EthSpec, Hot: ItemStore<E>, Cold: ItemStore<E>> {
     pub hot_db: Hot,
     /// LRU cache of deserialized blocks. Updated whenever a block is loaded.
     block_cache: Mutex<LruCache<Hash256, SignedBeaconBlock<E>>>,
+    /// LRU cache of replayed states.
+    state_cache: Mutex<LruCache<Slot, BeaconState<E>>>,
     /// Chain spec.
     pub(crate) spec: ChainSpec,
     /// Logger.
@@ -129,6 +131,7 @@ impl<E: EthSpec> HotColdDB<E, MemoryStore<E>, MemoryStore<E>> {
             cold_db: MemoryStore::open(),
             hot_db: MemoryStore::open(),
             block_cache: Mutex::new(LruCache::new(config.block_cache_size)),
+            state_cache: Mutex::new(LruCache::new(config.historic_state_cache_size)),
             config,
             spec,
             log,
@@ -162,6 +165,7 @@ impl<E: EthSpec> HotColdDB<E, LevelDB<E>, LevelDB<E>> {
             cold_db: LevelDB::open(cold_path)?,
             hot_db: LevelDB::open(hot_path)?,
             block_cache: Mutex::new(LruCache::new(config.block_cache_size)),
+            state_cache: Mutex::new(LruCache::new(config.historic_state_cache_size)),
             config,
             spec,
             log,
@@ -579,6 +583,9 @@ impl<E: EthSpec, Hot: ItemStore<E>, Cold: ItemStore<E>> HotColdDB<E, Hot, Cold>
     /// (which are frozen, and won't be deleted), or valid descendents of the finalized checkpoint
     /// (which will be deleted by this function but shouldn't be).
     pub fn delete_state(&self, state_root: &Hash256, slot: Slot) -> Result<(), Error> {
+        // Delete the state from the cache.
+        self.state_cache.lock().pop(&slot);
+
         // Delete the state summary.
         self.hot_db
             .key_delete(DBColumn::BeaconStateSummary.into(), state_root.as_bytes())?;
@@ -977,40 +984,69 @@ impl<E: EthSpec, Hot: ItemStore<E>, Cold: ItemStore<E>> HotColdDB<E, Hot, Cold>
 
     /// Load a frozen state that lies between restore points.
     fn load_cold_intermediate_state(&self, slot: Slot) -> Result<BeaconState<E>, Error> {
+        if let Some(state) = self.state_cache.lock().get(&slot) {
+            return Ok(state.clone());
+        }
+
         // 1. Load the restore points either side of the intermediate state.
         let low_restore_point_idx = slot.as_u64() / self.config.slots_per_restore_point;
         let high_restore_point_idx = low_restore_point_idx + 1;
 
+        // Use low restore point as the base state.
+        let mut low_slot: Slot =
+            Slot::new(low_restore_point_idx * self.config.slots_per_restore_point);
+        let mut low_state: Option<BeaconState<E>> = None;
+
+        // Try to get a more recent state from the cache to avoid massive blocks replay.
+        for (s, state) in self.state_cache.lock().iter() {
+            if s.as_u64() / self.config.slots_per_restore_point == low_restore_point_idx
+                && *s < slot
+                && low_slot < *s
+            {
+                low_slot = *s;
+                low_state = Some(state.clone());
+            }
+        }
+
+        // If low_state is still None, use load_restore_point_by_index to load the state.
+        if low_state.is_none() {
+            low_state = Some(self.load_restore_point_by_index(low_restore_point_idx)?);
+        }
+
         // Acquire the read lock, so that the split can't change while this is happening.
         let split = self.split.read_recursive();
 
-        let low_restore_point = self.load_restore_point_by_index(low_restore_point_idx)?;
         let high_restore_point = self.get_restore_point(high_restore_point_idx, &split)?;
 
-        // 2. Load the blocks from the high restore point back to the low restore point.
+        // 2. Load the blocks from the high restore point back to the low point.
         let blocks = self.load_blocks_to_replay(
-            low_restore_point.slot(),
+            low_slot,
             slot,
             self.get_high_restore_point_block_root(&high_restore_point, slot)?,
         )?;
 
-        // 3. Replay the blocks on top of the low restore point.
+        // 3. Replay the blocks on top of the low point.
         // Use a forwards state root iterator to avoid doing any tree hashing.
         // The state root of the high restore point should never be used, so is safely set to 0.
         let state_root_iter = self.forwards_state_roots_iterator_until(
-            low_restore_point.slot(),
+            low_slot,
             slot,
             || (high_restore_point, Hash256::zero()),
             &self.spec,
         )?;
 
-        self.replay_blocks(
-            low_restore_point,
+        let state = self.replay_blocks(
+            low_state.unwrap(),
             blocks,
             slot,
             Some(state_root_iter),
             StateRootStrategy::Accurate,
-        )
+        )?;
+
+        // If state is not error, put it in the cache.
+        self.state_cache.lock().put(slot, state.clone());
+
+        Ok(state)
     }
 
     /// Get the restore point with the given index, or if it is out of bounds, the split state.

book/src/api-lighthouse.md

@@ -455,6 +455,7 @@ curl "http://localhost:5052/lighthouse/database/info" | jq
   "config": {
     "slots_per_restore_point": 2048,
     "block_cache_size": 5,
+    "historic_state_cache_size": 1,
     "compact_on_init": false,
     "compact_on_prune": true
   },

book/src/database-migrations.md

@@ -92,6 +92,7 @@ curl "http://localhost:5052/lighthouse/database/info"
     "slots_per_restore_point": 8192,
     "slots_per_restore_point_set_explicitly": true,
     "block_cache_size": 5,
+    "historic_state_cache_size": 1,
     "compact_on_init": false,
     "compact_on_prune": true
   }

lighthouse/tests/beacon_node.rs

@@ -1634,6 +1634,13 @@ fn block_cache_size_flag() {
         .with_config(|config| assert_eq!(config.store.block_cache_size, 4_usize));
 }
 #[test]
+fn historic_state_cache_size_flag() {
+    CommandLineTest::new()
+        .flag("historic-state-cache-size", Some("4"))
+        .run_with_zero_port()
+        .with_config(|config| assert_eq!(config.store.historic_state_cache_size, 4_usize));
+}
+#[test]
 fn auto_compact_db_flag() {
     CommandLineTest::new()
         .flag("auto-compact-db", Some("false"))