[ETCM-103] Move hardcoded values to configuration

Author: KonradStaniec

src/main/resources/application.conf

@@ -392,6 +392,17 @@ mantis {
     # Sets max number of blocks that can be stored in queue to import on fetcher side
     # Warning! This setting affects ability to go back in case of branch resolution so it should not be too low
     max-fetcher-queue-size = 1000
+
+    # Expected size fo state sync bloom filter.
+    # Current Size of ETC state trie is aroud 150M Nodes, so 200M is set to have some reserve
+    # If the number of elements inserted into bloom filter would be significally higher that expected, then number
+    # of false positives would rise which would degrade performance of state sync
+    state-sync-bloomFilter-size = 200000000
+
+
+    # Max number of mpt nodes held in memory in state sync, before saving them into database
+    # 100k is around 60mb (each key-value pair has around 600bytes)
+    state-sync-persistBatch-size = 100000
   }
 
   pruning {

src/main/scala/io/iohk/ethereum/blockchain/sync/FastSync.scala

@@ -144,7 +144,8 @@ class FastSync(
       "state-downloader"
     )
     private val syncStateScheduler = context.actorOf(
-      SyncStateSchedulerActor.props(syncStateDownloader, SyncStateScheduler(blockchain)),
+      SyncStateSchedulerActor
+        .props(syncStateDownloader, SyncStateScheduler(blockchain, syncConfig.stateSyncBloomFilterSize), syncConfig),
       "state-scheduler"
     )
 

src/main/scala/io/iohk/ethereum/blockchain/sync/SyncStateScheduler.scala

@@ -279,15 +279,13 @@ object SyncStateScheduler {
     }
   }
 
-  val mptTrieSize = 200000000
-
-  def getEmptyFilter: BloomFilter[ByteString] = {
-    BloomFilter.create[ByteString](ByteStringFunnel, mptTrieSize)
+  def getEmptyFilter(expectedFilterSize: Int): BloomFilter[ByteString] = {
+    BloomFilter.create[ByteString](ByteStringFunnel, expectedFilterSize)
   }
   // TODO add method to load bloom filter after node restart. Perfect way to do it would be to expose Observable
   // in RocksDBDataSource which underneath would use RockDbIterator which would traverse whole namespace.
-  def apply(blockchain: Blockchain): SyncStateScheduler = {
-    new SyncStateScheduler(blockchain, getEmptyFilter)
+  def apply(blockchain: Blockchain, expectedBloomFilterSize: Int): SyncStateScheduler = {
+    new SyncStateScheduler(blockchain, getEmptyFilter(expectedBloomFilterSize))
   }
 
   final case class StateNodeRequest(

src/main/scala/io/iohk/ethereum/blockchain/sync/SyncStateSchedulerActor.scala

@@ -12,14 +12,14 @@ import io.iohk.ethereum.blockchain.sync.SyncStateSchedulerActor.{
   RestartRequested,
   StartSyncingTo,
   StateSyncFinished,
-  WaitingForNewTargetBlock,
-  maxMembatchSize
+  WaitingForNewTargetBlock
 }
 import io.iohk.ethereum.utils.ByteStringUtils
+import io.iohk.ethereum.utils.Config.SyncConfig
 
 import scala.concurrent.duration._
 
-class SyncStateSchedulerActor(downloader: ActorRef, sync: SyncStateScheduler)
+class SyncStateSchedulerActor(downloader: ActorRef, sync: SyncStateScheduler, syncConfig: SyncConfig)
     extends Actor
     with ActorLogging
     with Timers {
@@ -87,7 +87,7 @@ class SyncStateSchedulerActor(downloader: ActorRef, sync: SyncStateScheduler)
               downloader ! GetMissingNodes(missing)
             }
 
-            if (state2.memBatch.size >= maxMembatchSize) {
+            if (state2.memBatch.size >= syncConfig.stateSyncPersistBatchSize) {
               log.debug("Current membatch size is {}, persisting nodes to database", state2.memBatch.size)
               val state3 = sync.persistBatch(state2, targetBlock)
               context.become(
@@ -126,8 +126,8 @@ class SyncStateSchedulerActor(downloader: ActorRef, sync: SyncStateScheduler)
 }
 
 object SyncStateSchedulerActor {
-  def props(downloader: ActorRef, sync: SyncStateScheduler): Props = {
-    Props(new SyncStateSchedulerActor(downloader, sync))
+  def props(downloader: ActorRef, sync: SyncStateScheduler, syncConfig: SyncConfig): Props = {
+    Props(new SyncStateSchedulerActor(downloader, sync, syncConfig))
   }
 
   final case object PrintInfo
@@ -141,9 +141,6 @@ object SyncStateSchedulerActor {
 
   case object StateSyncFinished
 
-  // TODO determine this number of maybe it should be put to config
-  val maxMembatchSize = 100000
-
   case object RestartRequested
   case object WaitingForNewTargetBlock
 

src/main/scala/io/iohk/ethereum/utils/Config.scala

@@ -123,7 +123,9 @@ object Config {
       fastSyncBlockValidationN: Int,
       fastSyncBlockValidationX: Int,
       maxTargetDifference: Int,
-      maximumTargetUpdateFailures: Int
+      maximumTargetUpdateFailures: Int,
+      stateSyncBloomFilterSize: Int,
+      stateSyncPersistBatchSize: Int,
   )
 
   object SyncConfig {
@@ -162,7 +164,9 @@ object Config {
         fastSyncBlockValidationN = syncConfig.getInt("fast-sync-block-validation-n"),
         fastSyncBlockValidationX = syncConfig.getInt("fast-sync-block-validation-x"),
         maxTargetDifference = syncConfig.getInt("max-target-difference"),
-        maximumTargetUpdateFailures = syncConfig.getInt("maximum-target-update-failures")
+        maximumTargetUpdateFailures = syncConfig.getInt("maximum-target-update-failures"),
+        stateSyncBloomFilterSize = syncConfig.getInt("state-sync-bloomFilter-size"),
+        stateSyncPersistBatchSize = syncConfig.getInt("state-sync-persistBatch-size")
       )
     }
   }

src/test/resources/application.conf

@@ -118,6 +118,9 @@ mantis {
     min-peers-to-choose-pivot-block = 2
     peers-to-choose-pivot-block-margin = 1
     pivot-block-offset = 500
+    state-sync-bloomFilter-size = 20000
+
+    state-sync-persistBatch-size = 10000
   }
 
   keyStore {

src/test/scala/io/iohk/ethereum/blockchain/sync/StateSyncSpec.scala

@@ -176,7 +176,8 @@ class StateSyncSpec
     lazy val scheduler = system.actorOf(
       SyncStateSchedulerActor.props(
         downloader,
-        new SyncStateScheduler(buildBlockChain(), SyncStateScheduler.getEmptyFilter)
+        new SyncStateScheduler(buildBlockChain(), SyncStateScheduler.getEmptyFilter(syncConfig.stateSyncBloomFilterSize)),
+        syncConfig
       )
     )
   }

src/test/scala/io/iohk/ethereum/blockchain/sync/SyncSchedulerSpec.scala

@@ -222,6 +222,7 @@ class SyncSchedulerSpec extends AnyFlatSpec with Matchers with EitherValues with
       val freshBlockchain = BlockchainImpl(freshStorage.storages)
       new TrieProvider(freshBlockchain, blockchainConfig)
     }
+    val bloomFilterSize = 1000
 
     def buildScheduler(): (
         SyncStateScheduler,
@@ -230,7 +231,7 @@ class SyncSchedulerSpec extends AnyFlatSpec with Matchers with EitherValues with
     ) = {
       val freshStorage = getNewStorages
       val freshBlockchain = BlockchainImpl(freshStorage.storages)
-      (SyncStateScheduler(freshBlockchain), freshBlockchain, freshStorage)
+      (SyncStateScheduler(freshBlockchain, bloomFilterSize), freshBlockchain, freshStorage)
     }
 
     def exchangeSingleNode(

src/test/scala/io/iohk/ethereum/blockchain/sync/TestSyncConfig.scala

@@ -39,7 +39,10 @@ trait TestSyncConfig extends SyncConfigBuilder {
     fastSyncBlockValidationN = 2048,
     fastSyncBlockValidationX = 50,
     maxTargetDifference = 5,
-    maximumTargetUpdateFailures = 1
+    maximumTargetUpdateFailures = 1,
+    stateSyncBloomFilterSize = 1000,
+    stateSyncPersistBatchSize = 1000
+
   )
 
   override lazy val syncConfig: SyncConfig = defaultSyncConfig