feat(book): [prune] config section (#4328)

bin/reth/src/args/pruning_args.rs

@@ -3,7 +3,7 @@
 use clap::Args;
 use reth_config::config::PruneConfig;
 use reth_primitives::{
-    ChainSpec, ContractLogsPruneConfig, PruneMode, PruneModes, MINIMUM_PRUNING_DISTANCE,
+    ChainSpec, PruneMode, PruneModes, ReceiptsLogPruneConfig, MINIMUM_PRUNING_DISTANCE,
 };
 use std::sync::Arc;
 
@@ -35,7 +35,7 @@ impl PruningArgs {
                         .map(|contract| PruneMode::Before(contract.block)),
                     account_history: Some(PruneMode::Distance(MINIMUM_PRUNING_DISTANCE)),
                     storage_history: Some(PruneMode::Distance(MINIMUM_PRUNING_DISTANCE)),
-                    contract_logs_filter: ContractLogsPruneConfig(
+                    receipts_log_filter: ReceiptsLogPruneConfig(
                         _chain_spec
                             .deposit_contract
                             .as_ref()

book/run/config.md

@@ -27,6 +27,7 @@ The configuration file contains the following sections:
   - [`reputation_weights`](#reputation_weights)
   - [`backoff_durations`](#backoff_durations)
 - [`[sessions]`](#the-sessions-section)
+- [`[prune]`](#the-prune-section)
 
 ## The `[stages]` section
 
@@ -330,4 +331,56 @@ secs = 120
 nanos = 0
 ```
 
+## The `[prune]` section
+
+The prune section configures the pruning configuration.
+
+You can configure the pruning of different parts of the data independently of others.
+For any unspecified parts, the default setting is no pruning.
+
+### Default config
+
+No pruning, run as archive node.
+
+### Example of the custom pruning configuration
+
+This configuration will:
+- Run pruning every 5 blocks
+- Continuously prune all transaction senders, account history and storage history before the block `head-128`, i.e. keep the data for the last 129 blocks
+- Prune all receipts before the block 1920000, i.e. keep receipts from the block 1920000
+
+```toml
+[prune]
+# Minimum pruning interval measured in blocks
+block_interval = 5
+
+[prune.parts]
+# Sender Recovery pruning configuration
+sender_recovery = { distance = 128 } # Prune all transaction senders before the block `head-128`, i.e. keep transaction senders for the last 129 blocks
+
+# Transaction Lookup pruning configuration
+transaction_lookup = "full" # Prune all TxNumber => TxHash mappings
+
+# Receipts pruning configuration. This setting overrides `receipts_log_filter`.
+receipts = { before = 1920000 } # Prune all receipts from transactions before the block 1920000, i.e. keep receipts from the block 1920000
+
+# Account History pruning configuration
+account_history = { distance = 128 } # Prune all historical account states before the block `head-128`
+
+# Storage History pruning configuration
+storage_history = { distance = 128 } # Prune all historical storage states before the block `head-128`
+```
+
+We can also prune receipts more granular, using the logs filtering:
+```toml
+# Receipts pruning configuration by retaining only those receipts that contain logs emitted
+# by the specified addresses, discarding all others. This setting is overridden by `receipts`.
+[prune.parts.receipts_log_filter]
+# Prune all receipts, leaving only those which:
+# - Contain logs from address `0x7ea2be2df7ba6e54b1a9c70676f668455e329d29`, starting from the block 17000000
+# - Contain logs from address `0xdac17f958d2ee523a2206206994597c13d831ec7` in the last 1001 blocks
+"0x7ea2be2df7ba6e54b1a9c70676f668455e329d29" = { before = 17000000 }
+"0xdac17f958d2ee523a2206206994597c13d831ec7" = { distance = 1000 }
+```
+
 [TOML]: https://toml.io/

crates/primitives/src/lib.rs

@@ -81,7 +81,7 @@ pub use net::{
 };
 pub use peer::{PeerId, WithPeerId};
 pub use prune::{
-    ContractLogsPruneConfig, PruneCheckpoint, PruneMode, PruneModes, PrunePart, PrunePartError,
+    PruneCheckpoint, PruneMode, PruneModes, PrunePart, PrunePartError, ReceiptsLogPruneConfig,
     MINIMUM_PRUNING_DISTANCE,
 };
 pub use receipt::{Receipt, ReceiptWithBloom, ReceiptWithBloomRef};

crates/primitives/src/prune/mod.rs

@@ -13,9 +13,9 @@ pub use target::{PruneModes, MINIMUM_PRUNING_DISTANCE};
 
 /// Configuration for pruning receipts not associated with logs emitted by the specified contracts.
 #[derive(Debug, Clone, PartialEq, Eq, Default, Serialize, Deserialize)]
-pub struct ContractLogsPruneConfig(pub BTreeMap<Address, PruneMode>);
+pub struct ReceiptsLogPruneConfig(pub BTreeMap<Address, PruneMode>);
 
-impl ContractLogsPruneConfig {
+impl ReceiptsLogPruneConfig {
     /// Checks if the configuration is empty
     pub fn is_empty(&self) -> bool {
         self.0.is_empty()

crates/primitives/src/prune/target.rs

@@ -1,6 +1,6 @@
 use crate::{
     prune::PrunePartError, serde_helper::deserialize_opt_prune_mode_with_min_blocks, BlockNumber,
-    ContractLogsPruneConfig, PruneMode, PrunePart,
+    PruneMode, PrunePart, ReceiptsLogPruneConfig,
 };
 use paste::paste;
 use serde::{Deserialize, Serialize};
@@ -23,8 +23,8 @@ pub struct PruneModes {
     /// Transaction Lookup pruning configuration.
     #[serde(skip_serializing_if = "Option::is_none")]
     pub transaction_lookup: Option<PruneMode>,
-    /// Configuration for pruning of receipts. This setting overrides
-    /// `PruneModes::contract_logs_filter` and offers improved performance.
+    /// Receipts pruning configuration. This setting overrides `receipts_log_filter`
+    /// and offers improved performance.
     #[serde(
         skip_serializing_if = "Option::is_none",
         deserialize_with = "deserialize_opt_prune_mode_with_min_blocks::<64, _>"
@@ -42,12 +42,12 @@ pub struct PruneModes {
         deserialize_with = "deserialize_opt_prune_mode_with_min_blocks::<64, _>"
     )]
     pub storage_history: Option<PruneMode>,
-    /// Retains only those receipts that contain logs emitted by the specified addresses,
-    /// discarding all others. Note that this setting is overridden by `PruneModes::receipts`.
+    /// Receipts pruning configuration by retaining only those receipts that contain logs emitted
+    /// by the specified addresses, discarding others. This setting is overridden by `receipts`.
     ///
     /// The [`BlockNumber`] represents the starting block from which point onwards the receipts are
     /// preserved.
-    pub contract_logs_filter: ContractLogsPruneConfig,
+    pub receipts_log_filter: ReceiptsLogPruneConfig,
 }
 
 macro_rules! impl_prune_parts {
@@ -90,7 +90,7 @@ macro_rules! impl_prune_parts {
                 $(
                     $part: Some(PruneMode::Full),
                 )+
-                contract_logs_filter: Default::default()
+                receipts_log_filter: Default::default()
             }
         }
 

crates/prune/src/pruner.rs

@@ -103,7 +103,7 @@ impl<DB: Database> Pruner<DB> {
                 .record(part_start.elapsed())
         }
 
-        if !self.modes.contract_logs_filter.is_empty() {
+        if !self.modes.receipts_log_filter.is_empty() {
             let part_start = Instant::now();
             self.prune_receipts_by_logs(&provider, tip_block_number)?;
             self.metrics
@@ -305,7 +305,7 @@ impl<DB: Database> Pruner<DB> {
             .map(|checkpoint| checkpoint.block_number);
 
         let address_filter =
-            self.modes.contract_logs_filter.group_by_block(tip_block_number, pruned)?;
+            self.modes.receipts_log_filter.group_by_block(tip_block_number, pruned)?;
 
         // Splits all transactions in different block ranges. Each block range will have its own
         // filter address list and will check it while going through the table
@@ -411,7 +411,7 @@ impl<DB: Database> Pruner<DB> {
         // one using `get_next_tx_num_range_from_checkpoint`.
         let checkpoint_block = self
             .modes
-            .contract_logs_filter
+            .receipts_log_filter
             .lowest_block_with_distance(tip_block_number, pruned)?
             .unwrap_or(to_block);
 

crates/storage/provider/src/post_state/mod.rs

@@ -661,7 +661,7 @@ impl PostState {
 
             let contract_log_pruner = self
                 .prune_modes
-                .contract_logs_filter
+                .receipts_log_filter
                 .group_by_block(tip, None)
                 .map_err(|e| Error::Custom(e.to_string()))?;