From c8df172063db916712955fe6cad94d71523b3260 Mon Sep 17 00:00:00 2001
From: Alexey Shekhirin <a.shekhirin@gmail.com>
Date: Tue, 5 Mar 2024 19:20:44 +0000
Subject: [PATCH] fix(stages): calculate checkpoints from static files (#6973)

---
 crates/stages/src/stages/bodies.rs          | 6 +++++-
 crates/stages/src/stages/sender_recovery.rs | 5 ++++-
 crates/stages/src/stages/tx_lookup.rs       | 5 ++++-
 3 files changed, 13 insertions(+), 3 deletions(-)

diff --git a/crates/stages/src/stages/bodies.rs b/crates/stages/src/stages/bodies.rs
index 459eace720..312c465798 100644
--- a/crates/stages/src/stages/bodies.rs
+++ b/crates/stages/src/stages/bodies.rs
@@ -343,7 +343,11 @@ fn stage_checkpoint<DB: Database>(
 ) -> ProviderResult<EntitiesCheckpoint> {
     Ok(EntitiesCheckpoint {
         processed: provider.count_entries::<tables::BlockBodyIndices>()? as u64,
-        total: (provider.count_entries::<tables::Headers>()? as u64).saturating_sub(1),
+        // Count only static files entries. If we count the database entries too, we may have
+        // duplicates. We're sure that the static files have all entries that database has,
+        // because we run the `StaticFileProducer` before starting the pipeline.
+        total: (provider.static_file_provider().count_entries::<tables::Headers>()? as u64)
+            .saturating_sub(1),
     })
 }
 
diff --git a/crates/stages/src/stages/sender_recovery.rs b/crates/stages/src/stages/sender_recovery.rs
index 79e4263661..6ec8f71cb3 100644
--- a/crates/stages/src/stages/sender_recovery.rs
+++ b/crates/stages/src/stages/sender_recovery.rs
@@ -221,7 +221,10 @@ fn stage_checkpoint<DB: Database>(
         // matching the actual number of processed transactions. To fix that, we add the
         // number of pruned `TransactionSenders` entries.
         processed: provider.count_entries::<tables::TransactionSenders>()? as u64 + pruned_entries,
-        total: provider.count_entries::<tables::Transactions>()? as u64,
+        // Count only static files entries. If we count the database entries too, we may have
+        // duplicates. We're sure that the static files have all entries that database has,
+        // because we run the `StaticFileProducer` before starting the pipeline.
+        total: provider.static_file_provider().count_entries::<tables::Transactions>()? as u64,
     })
 }
 
diff --git a/crates/stages/src/stages/tx_lookup.rs b/crates/stages/src/stages/tx_lookup.rs
index ac68997541..ef1f51eefa 100644
--- a/crates/stages/src/stages/tx_lookup.rs
+++ b/crates/stages/src/stages/tx_lookup.rs
@@ -219,7 +219,10 @@ fn stage_checkpoint<DB: Database>(
         // number of pruned `TransactionHashNumbers` entries.
         processed: provider.count_entries::<tables::TransactionHashNumbers>()? as u64 +
             pruned_entries,
-        total: provider.count_entries::<tables::Transactions>()? as u64,
+        // Count only static files entries. If we count the database entries too, we may have
+        // duplicates. We're sure that the static files have all entries that database has,
+        // because we run the `StaticFileProducer` before starting the pipeline.
+        total: provider.static_file_provider().count_entries::<tables::Transactions>()? as u64,
     })
 }