From d9bff54d570aff49ce7ee78ada9a4eed6fb21a25 Mon Sep 17 00:00:00 2001
From: Jacek Sieka <jacek@status.im>
Date: Sat, 24 Jun 2023 16:14:28 +0200
Subject: [PATCH] don't check legacy tables for pruning (#5116)

These tables can't be deleted from (read-only) and would be too slow to
delete from anyway due to the inefficient storage format in use.

* slow down startup clearing too
* remove unused del function
---
 beacon_chain/beacon_chain_db.nim              | 15 ++++---------
 .../consensus_object_pools/blockchain_dag.nim | 21 ++++++++++++-------
 2 files changed, 18 insertions(+), 18 deletions(-)

diff --git a/beacon_chain/beacon_chain_db.nim b/beacon_chain/beacon_chain_db.nim
index 9652bbb03..97b913f48 100644
--- a/beacon_chain/beacon_chain_db.nim
+++ b/beacon_chain/beacon_chain_db.nim
@@ -885,18 +885,11 @@ proc delBlock*(db: BeaconChainDB, fork: ConsensusFork, key: Eth2Digest): bool =
 proc delState*(db: BeaconChainDB, fork: ConsensusFork, key: Eth2Digest) =
   discard db.statesNoVal[fork].del(key.data).expectDb()
 
-proc clearBlocks*(db: BeaconChainDB, fork: ConsensusFork) =
-  discard db.blocks[fork].clear().expectDb()
+proc clearBlocks*(db: BeaconChainDB, fork: ConsensusFork): bool =
+  db.blocks[fork].clear().expectDb()
 
-proc clearStates*(db: BeaconChainDB, fork: ConsensusFork) =
-  discard db.statesNoVal[fork].clear().expectDb()
-
-proc delKeyValue*(db: BeaconChainDB, key: array[1, byte]) =
-  discard db.keyValues.del(key).expectDb()
-  discard db.v0.backend.del(key).expectDb()
-
-proc delKeyValue*(db: BeaconChainDB, key: DbKeyKind) =
-  db.delKeyValue(subkey(key))
+proc clearStates*(db: BeaconChainDB, fork: ConsensusFork): bool =
+  db.statesNoVal[fork].clear().expectDb()
 
 proc delStateRoot*(db: BeaconChainDB, root: Eth2Digest, slot: Slot) =
   discard db.stateRoots.del(stateRootKey(root, slot)).expectDb()
diff --git a/beacon_chain/consensus_object_pools/blockchain_dag.nim b/beacon_chain/consensus_object_pools/blockchain_dag.nim
index 9b2081f40..7d8f89639 100644
--- a/beacon_chain/consensus_object_pools/blockchain_dag.nim
+++ b/beacon_chain/consensus_object_pools/blockchain_dag.nim
@@ -683,13 +683,13 @@ proc getState(
 
 proc containsState*(
     db: BeaconChainDB, cfg: RuntimeConfig, block_root: Eth2Digest,
-    slots: Slice[Slot]): bool =
+    slots: Slice[Slot], legacy = true): bool =
   var slot = slots.b
   while slot >= slots.a:
     let state_root = db.getStateRoot(block_root, slot)
     if state_root.isSome() and
         db.containsState(
-          cfg.consensusForkAtEpoch(slot.epoch), state_root.get()):
+          cfg.consensusForkAtEpoch(slot.epoch), state_root.get(), legacy):
       return true
 
     if slot == slots.a: # avoid underflow at genesis
@@ -2192,7 +2192,10 @@ proc pruneHistory*(dag: ChainDAGRef, startup = false) =
       var first = true
       while cur.isSome():
         let bs = cur.get()
-        if dag.db.containsState(dag.cfg, bs.bid.root, bs.slot..bs.slot):
+        # We don't delete legacy states because the legacy database is openend
+        # in read-only and slow to delete from due to its sub-optimal structure
+        if dag.db.containsState(
+            dag.cfg, bs.bid.root, bs.slot..bs.slot, legacy = first):
           if first:
             # We leave the state on the prune horizon intact and update the tail
             # to point to this state, indicating the new point in time from
@@ -2252,17 +2255,21 @@ proc pruneHistory*(dag: ChainDAGRef, startup = false) =
       # Once during start, we'll clear all "old fork" data - this ensures we get
       # rid of any leftover junk in the tables - we do so after linear pruning
       # so as to "mostly" clean up the phase0 tables as well (which cannot be
-      # pruned easily by fork)
+      # pruned easily by fork) - one fork at a time, so as not to take too long
 
       let stateFork = dag.cfg.consensusForkAtEpoch(dag.tail.slot.epoch)
+      var clearedStates = false
       if stateFork > ConsensusFork.Phase0:
         for fork in ConsensusFork.Phase0..<stateFork:
-          dag.db.clearStates(fork)
+          if dag.db.clearStates(fork):
+            clearedStates = true
+            break
 
       let blockFork = dag.cfg.consensusForkAtEpoch(blockHorizon.epoch)
-      if blockFork > ConsensusFork.Phase0:
+      if not clearedStates and blockFork > ConsensusFork.Phase0:
         for fork in ConsensusFork.Phase0..<blockFork:
-          dag.db.clearBlocks(fork)
+          if dag.db.clearBlocks(fork):
+            break
 
 proc loadExecutionBlockHash*(dag: ChainDAGRef, bid: BlockId): Eth2Digest =
   if dag.cfg.consensusForkAtEpoch(bid.slot.epoch) < ConsensusFork.Bellatrix: