Use a smaller thread pool during pre-cache phase of GC to reduce memory pressure (#205)

2 files changed, 12 insertions, 7 deletions

diff --git a/CHANGELOG.md b/CHANGELOG.md
index fc5b227ba..85b5f1022 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -6,6 +6,7 @@
   - Separate page for inspecting server stats
 - Improvement: When batch-fetching cache values/attachments, only use block memory-buffering for items that may be memcached to reduce disk I/O
 - Improvement: `zen project-drop` command defaults to dry-run to prevent accidental drop and provides information on how to actually perform the drop
+- Improvement: Use a smaller thread pool during pre-cache phase of GC to reduce memory pressure
 - Bugfix: Parse filenames as utf8 string when mirroring files inside an oplog using `zen oplog-mirror` and `zen vfs`
 - Bugfix: Properly initialize stats for file/jupiter/zen remote project store - fixes stats log output when exporting and importing oplogs
 - Cleanup: Removed GCv1 implementation
diff --git a/src/zenstore/gc.cpp b/src/zenstore/gc.cpp
index 981ba15cb..be8fc0148 100644
--- a/src/zenstore/gc.cpp
+++ b/src/zenstore/gc.cpp
@@ -632,7 +632,8 @@ GcManager::CollectGarbage(const GcSettings& Settings)
 
 		std::unordered_map<std::unique_ptr<GcStoreCompactor>, GcCompactStoreStats*> StoreCompactors;
 		RwLock																		StoreCompactorsLock;
-		WorkerThreadPool& ThreadPool = Settings.SingleThread ? GetSyncWorkerPool() : GetMediumWorkerPool(EWorkloadType::Background);
+		WorkerThreadPool&															PreCachePhaseThreadPool =
+			  Settings.SingleThread ? GetSyncWorkerPool() : GetSmallWorkerPool(EWorkloadType::Background);
 
 		ZEN_INFO("GCV2: Removing expired data from {} referencers", m_GcReferencers.size());
 		if (!m_GcReferencers.empty())
@@ -662,7 +663,7 @@ GcManager::CollectGarbage(const GcSettings& Settings)
 					GcReferencer*							   Owner = m_GcReferencers[Index];
 					std::pair<std::string, GcReferencerStats>* Stats = &Result.ReferencerStats[Index];
 					WorkLeft.AddCount(1);
-					ThreadPool.ScheduleWork([this, &Ctx, &WorkLeft, Owner, Stats, &StoreCompactorsLock, &StoreCompactors]() {
+					PreCachePhaseThreadPool.ScheduleWork([this, &Ctx, &WorkLeft, Owner, Stats, &StoreCompactorsLock, &StoreCompactors]() {
 						auto _ = MakeGuard([&WorkLeft]() { WorkLeft.CountDown(); });
 						try
 						{
@@ -728,7 +729,7 @@ GcManager::CollectGarbage(const GcSettings& Settings)
 						GcReferenceStore*							   ReferenceStore = m_GcReferenceStores[Index];
 						std::pair<std::string, GcReferenceStoreStats>* Stats		  = &Result.ReferenceStoreStats[Index];
 						WorkLeft.AddCount(1);
-						ThreadPool.ScheduleWork(
+						PreCachePhaseThreadPool.ScheduleWork(
 							[this, &Ctx, ReferenceStore, Stats, Index, &WorkLeft, &ReferencePrunersLock, &ReferencePruners]() {
 								auto _ = MakeGuard([&WorkLeft]() { WorkLeft.CountDown(); });
 								try
@@ -799,7 +800,7 @@ GcManager::CollectGarbage(const GcSettings& Settings)
 							GcReferencer*							   Referencer = m_GcReferencers[Index];
 							std::pair<std::string, GcReferencerStats>* Stats	  = &Result.ReferencerStats[Index];
 							WorkLeft.AddCount(1);
-							ThreadPool.ScheduleWork(
+							PreCachePhaseThreadPool.ScheduleWork(
 								[this, &Ctx, &WorkLeft, Referencer, Index, Stats, &ReferenceCheckersLock, &ReferenceCheckers]() {
 									auto _ = MakeGuard([&WorkLeft]() { WorkLeft.CountDown(); });
 									// The Referencer will create a reference checker that guarantees that the references do not change as
@@ -873,7 +874,7 @@ GcManager::CollectGarbage(const GcSettings& Settings)
 								size_t									   Index   = It.second;
 								std::pair<std::string, GcReferencerStats>* Stats   = &Result.ReferencerStats[Index];
 								WorkLeft.AddCount(1);
-								ThreadPool.ScheduleWork([this, &Ctx, Checker, Index, Stats, &WorkLeft]() {
+								PreCachePhaseThreadPool.ScheduleWork([this, &Ctx, Checker, Index, Stats, &WorkLeft]() {
 									auto _ = MakeGuard([&WorkLeft]() { WorkLeft.CountDown(); });
 									try
 									{
@@ -893,6 +894,9 @@ GcManager::CollectGarbage(const GcSettings& Settings)
 					}
 				}
 
+				WorkerThreadPool& LockedPhaseThreadPool =
+					Settings.SingleThread ? GetSyncWorkerPool() : GetMediumWorkerPool(EWorkloadType::Background);
+
 				std::vector<RwLock::SharedLockScope> LockerScopes;
 				SCOPED_TIMER(uint64_t ElapsedMS = Timer.GetElapsedTimeMs(); Result.WriteBlockMS = std::chrono::milliseconds(ElapsedMS);
 							 ZEN_INFO("GCV2: Writes blocked for {}", NiceTimeSpanMs(ElapsedMS)));
@@ -957,7 +961,7 @@ GcManager::CollectGarbage(const GcSettings& Settings)
 									size_t									   Index   = It.second;
 									std::pair<std::string, GcReferencerStats>* Stats   = &Result.ReferencerStats[Index];
 									WorkLeft.AddCount(1);
-									ThreadPool.ScheduleWork([this, &Ctx, Checker, Index, Stats, &WorkLeft]() {
+									LockedPhaseThreadPool.ScheduleWork([this, &Ctx, Checker, Index, Stats, &WorkLeft]() {
 										auto _ = MakeGuard([&WorkLeft]() { WorkLeft.CountDown(); });
 										try
 										{
@@ -1030,7 +1034,7 @@ GcManager::CollectGarbage(const GcSettings& Settings)
 								size_t				   Index  = It.first;
 								GcReferenceStoreStats* Stats  = &Result.ReferenceStoreStats[Index].second;
 								WorkLeft.AddCount(1);
-								ThreadPool.ScheduleWork(
+								LockedPhaseThreadPool.ScheduleWork(
 									[this, &Ctx, Pruner, Stats, &WorkLeft, &GetUnusedReferences, &StoreCompactorsLock, &StoreCompactors]() {
 										auto _ = MakeGuard([&WorkLeft]() { WorkLeft.CountDown(); });
 										// Go through all the ReferenceCheckers to see if the list of Cids the collector selected are