diff --git a/nano/core_test/confirmation_height.cpp b/nano/core_test/confirmation_height.cpp
index 9bae54bd..44597b14 100644
--- a/nano/core_test/confirmation_height.cpp
+++ b/nano/core_test/confirmation_height.cpp
@@ -1574,3 +1574,52 @@ TEST (confirmation_height, election_winner_details_clearing_node_process_confirm
 	node->process_confirmed (election, 1000000);
 	ASSERT_EQ (0, node->active.election_winner_details_size ());
 }
+
+TEST (confirmation_height, unbounded_block_cache_iteration)
+{
+	nano::logger_mt logger;
+	auto path (nano::unique_path ());
+	nano::mdb_store store (logger, path);
+	ASSERT_TRUE (!store.init_error ());
+	nano::genesis genesis;
+	nano::stat stats;
+	nano::ledger ledger (store, stats);
+	nano::write_database_queue write_database_queue;
+	boost::latch initialized_latch{ 0 };
+	nano::work_pool pool (std::numeric_limits<unsigned>::max ());
+	nano::keypair key1;
+	auto send = std::make_shared<nano::send_block> (genesis.hash (), key1.pub, nano::genesis_amount - nano::Gxrb_ratio, nano::test_genesis_key.prv, nano::test_genesis_key.pub, *pool.generate (genesis.hash ()));
+	auto send1 = std::make_shared<nano::send_block> (send->hash (), key1.pub, nano::genesis_amount - nano::Gxrb_ratio * 2, nano::test_genesis_key.prv, nano::test_genesis_key.pub, *pool.generate (send->hash ()));
+	{
+		auto transaction (store.tx_begin_write ());
+		store.initialize (transaction, genesis, ledger.cache);
+		ASSERT_EQ (nano::process_result::progress, ledger.process (transaction, *send).code);
+		ASSERT_EQ (nano::process_result::progress, ledger.process (transaction, *send1).code);
+	}
+
+	nano::confirmation_height_processor confirmation_height_processor (ledger, write_database_queue, 10ms, logger, initialized_latch, nano::confirmation_height_mode::unbounded);
+	nano::timer<> timer;
+	timer.start ();
+	{
+		// Prevent conf height processor doing any writes, so that we can query is_processing_block correctly
+		auto write_guard = write_database_queue.wait (nano::writer::testing);
+		// Add the frontier block
+		confirmation_height_processor.add (send1->hash ());
+
+		// The most uncemented block (previous block) should be seen as processing by the unbounded processor
+		while (!confirmation_height_processor.is_processing_block (send->hash ()))
+		{
+			ASSERT_LT (timer.since_start (), 10s);
+		}
+	}
+
+	// Wait until the current block is finished processing
+	while (!confirmation_height_processor.current ().is_zero ())
+	{
+		ASSERT_LT (timer.since_start (), 10s);
+	}
+
+	ASSERT_EQ (2, stats.count (nano::stat::type::confirmation_height, nano::stat::detail::blocks_confirmed, nano::stat::dir::in));
+	ASSERT_EQ (2, stats.count (nano::stat::type::confirmation_height, nano::stat::detail::blocks_confirmed_unbounded, nano::stat::dir::in));
+	ASSERT_EQ (3, ledger.cache.cemented_count);
+}
diff --git a/nano/node/active_transactions.cpp b/nano/node/active_transactions.cpp
index 0636224b..b0cebefb 100644
--- a/nano/node/active_transactions.cpp
+++ b/nano/node/active_transactions.cpp
@@ -124,7 +124,7 @@ void nano::active_transactions::block_cemented_callback (std::shared_ptr<nano::b
 	auto transaction = node.store.tx_begin_read ();
 
 	boost::optional<nano::election_status_type> election_status_type;
-	if (!confirmation_height_processor.is_processing_block (block_a->hash ()))
+	if (!confirmation_height_processor.is_processing_added_block (block_a->hash ()))
 	{
 		election_status_type = confirm_block (transaction, block_a);
 	}
diff --git a/nano/node/confirmation_height_processor.cpp b/nano/node/confirmation_height_processor.cpp
index d05ffeab..c37f6639 100644
--- a/nano/node/confirmation_height_processor.cpp
+++ b/nano/node/confirmation_height_processor.cpp
@@ -209,19 +209,24 @@ std::unique_ptr<nano::container_info_component> nano::collect_container_info (co
 	return composite;
 }
 
-size_t nano::confirmation_height_processor::awaiting_processing_size ()
+size_t nano::confirmation_height_processor::awaiting_processing_size () const
 {
 	nano::lock_guard<std::mutex> guard (mutex);
 	return awaiting_processing.size ();
 }
 
-bool nano::confirmation_height_processor::is_processing_block (nano::block_hash const & hash_a)
+bool nano::confirmation_height_processor::is_processing_added_block (nano::block_hash const & hash_a) const
 {
 	nano::lock_guard<std::mutex> guard (mutex);
 	return original_hashes_pending.count (hash_a) > 0 || awaiting_processing.get<tag_hash> ().count (hash_a) > 0;
 }
 
-nano::block_hash nano::confirmation_height_processor::current ()
+bool nano::confirmation_height_processor::is_processing_block (nano::block_hash const & hash_a) const
+{
+	return is_processing_added_block (hash_a) || unbounded_processor.has_iterated_over_block (hash_a);
+}
+
+nano::block_hash nano::confirmation_height_processor::current () const
 {
 	nano::lock_guard<std::mutex> lk (mutex);
 	return original_hash;
diff --git a/nano/node/confirmation_height_processor.hpp b/nano/node/confirmation_height_processor.hpp
index 60270a10..a0cfb6ec 100644
--- a/nano/node/confirmation_height_processor.hpp
+++ b/nano/node/confirmation_height_processor.hpp
@@ -37,15 +37,16 @@ public:
 	void stop ();
 	void add (nano::block_hash const & hash_a);
 	void run (confirmation_height_mode);
-	size_t awaiting_processing_size ();
-	bool is_processing_block (nano::block_hash const &);
-	nano::block_hash current ();
+	size_t awaiting_processing_size () const;
+	bool is_processing_added_block (nano::block_hash const & hash_a) const;
+	bool is_processing_block (nano::block_hash const &) const;
+	nano::block_hash current () const;
 
 	void add_cemented_observer (std::function<void(std::shared_ptr<nano::block>)> const &);
 	void add_block_already_cemented_observer (std::function<void(nano::block_hash const &)> const &);
 
 private:
-	std::mutex mutex;
+	mutable std::mutex mutex;
 	// Hashes which have been added to the confirmation height processor, but not yet processed
 	// clang-format off
 	class tag_sequence {};
diff --git a/nano/node/confirmation_height_unbounded.cpp b/nano/node/confirmation_height_unbounded.cpp
index 9cd29f02..612407d1 100644
--- a/nano/node/confirmation_height_unbounded.cpp
+++ b/nano/node/confirmation_height_unbounded.cpp
@@ -379,8 +379,9 @@ void nano::confirmation_height_unbounded::cement_blocks (nano::write_guard & sco
 				// Reverse it so that the callbacks start from the lowest newly cemented block and move upwards
 				std::reverse (pending.block_callback_data.begin (), pending.block_callback_data.end ());
 
+				nano::lock_guard<std::mutex> guard (block_cache_mutex);
 				std::transform (pending.block_callback_data.begin (), pending.block_callback_data.end (), std::back_inserter (cemented_blocks), [& block_cache = block_cache](auto const & hash_a) {
-					debug_assert (block_cache.find (hash_a) != block_cache.end ());
+					debug_assert (block_cache.count (hash_a) == 1);
 					return block_cache.at (hash_a);
 				});
 			}
@@ -413,6 +414,7 @@ void nano::confirmation_height_unbounded::cement_blocks (nano::write_guard & sco
 
 std::shared_ptr<nano::block> nano::confirmation_height_unbounded::get_block_and_sideband (nano::block_hash const & hash_a, nano::transaction const & transaction_a)
 {
+	nano::lock_guard<std::mutex> guard (block_cache_mutex);
 	auto block_cache_it = block_cache.find (hash_a);
 	if (block_cache_it != block_cache.cend ())
 	{
@@ -422,7 +424,6 @@ std::shared_ptr<nano::block> nano::confirmation_height_unbounded::get_block_and_
 	{
 		auto block (ledger.store.block_get (transaction_a, hash_a));
 		block_cache.emplace (hash_a, block);
-		++block_cache_size;
 		return block;
 	}
 }
@@ -440,8 +441,22 @@ void nano::confirmation_height_unbounded::clear_process_vars ()
 	confirmed_iterated_pairs_size = 0;
 	implicit_receive_cemented_mapping.clear ();
 	implicit_receive_cemented_mapping_size = 0;
-	block_cache.clear ();
-	block_cache_size = 0;
+	{
+		nano::lock_guard<std::mutex> guard (block_cache_mutex);
+		block_cache.clear ();
+	}
+}
+
+bool nano::confirmation_height_unbounded::has_iterated_over_block (nano::block_hash const & hash_a) const
+{
+	nano::lock_guard<std::mutex> guard (block_cache_mutex);
+	return block_cache.count (hash_a) == 1;
+}
+
+uint64_t nano::confirmation_height_unbounded::block_cache_size () const
+{
+	nano::lock_guard<std::mutex> guard (block_cache_mutex);
+	return block_cache.size ();
 }
 
 nano::confirmation_height_unbounded::conf_height_details::conf_height_details (nano::account const & account_a, nano::block_hash const & hash_a, uint64_t height_a, uint64_t num_blocks_confirmed_a, std::vector<nano::block_hash> const & block_callback_data_a) :
@@ -471,6 +486,6 @@ std::unique_ptr<nano::container_info_component> nano::collect_container_info (co
 	composite->add_component (std::make_unique<container_info_leaf> (container_info{ "confirmed_iterated_pairs", confirmation_height_unbounded.confirmed_iterated_pairs_size, sizeof (decltype (confirmation_height_unbounded.confirmed_iterated_pairs)::value_type) }));
 	composite->add_component (std::make_unique<container_info_leaf> (container_info{ "pending_writes", confirmation_height_unbounded.pending_writes_size, sizeof (decltype (confirmation_height_unbounded.pending_writes)::value_type) }));
 	composite->add_component (std::make_unique<container_info_leaf> (container_info{ "implicit_receive_cemented_mapping", confirmation_height_unbounded.implicit_receive_cemented_mapping_size, sizeof (decltype (confirmation_height_unbounded.implicit_receive_cemented_mapping)::value_type) }));
-	composite->add_component (std::make_unique<container_info_leaf> (container_info{ "block_cache", confirmation_height_unbounded.block_cache_size, sizeof (decltype (confirmation_height_unbounded.block_cache)::value_type) }));
+	composite->add_component (std::make_unique<container_info_leaf> (container_info{ "block_cache", confirmation_height_unbounded.block_cache_size (), sizeof (decltype (confirmation_height_unbounded.block_cache)::value_type) }));
 	return composite;
 }
diff --git a/nano/node/confirmation_height_unbounded.hpp b/nano/node/confirmation_height_unbounded.hpp
index 8a8d1aa1..20cfc544 100644
--- a/nano/node/confirmation_height_unbounded.hpp
+++ b/nano/node/confirmation_height_unbounded.hpp
@@ -23,6 +23,7 @@ public:
 	void clear_process_vars ();
 	void process ();
 	void cement_blocks (nano::write_guard &);
+	bool has_iterated_over_block (nano::block_hash const &) const;
 
 private:
 	class confirmed_iterated_pair
@@ -62,14 +63,16 @@ private:
 	// This allows the load and stores to use relaxed atomic memory ordering.
 	std::unordered_map<account, confirmed_iterated_pair> confirmed_iterated_pairs;
 	nano::relaxed_atomic_integral<uint64_t> confirmed_iterated_pairs_size{ 0 };
-	std::unordered_map<nano::block_hash, std::shared_ptr<nano::block>> block_cache;
-	nano::relaxed_atomic_integral<uint64_t> block_cache_size{ 0 };
 	std::shared_ptr<nano::block> get_block_and_sideband (nano::block_hash const &, nano::transaction const &);
 	std::deque<conf_height_details> pending_writes;
 	nano::relaxed_atomic_integral<uint64_t> pending_writes_size{ 0 };
 	std::unordered_map<nano::block_hash, std::weak_ptr<conf_height_details>> implicit_receive_cemented_mapping;
 	nano::relaxed_atomic_integral<uint64_t> implicit_receive_cemented_mapping_size{ 0 };
 
+	mutable std::mutex block_cache_mutex;
+	std::unordered_map<nano::block_hash, std::shared_ptr<nano::block>> block_cache;
+	uint64_t block_cache_size () const;
+
 	nano::timer<std::chrono::milliseconds> timer;
 
 	class preparation_data final