Skip to content

support memory pools configurations & tune spilling settings #7510

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 1 commit into from
Aug 7, 2024
Merged

support memory pools configurations & tune spilling settings #7510

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
11 changes: 3 additions & 8 deletions ydb/core/kqp/compute_actor/kqp_compute_actor_factory.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -14,15 +14,13 @@ struct TMemoryQuotaManager : public NYql::NDq::TGuaranteeQuotaManager {
, std::shared_ptr<IKqpNodeState> state
, TIntrusivePtr<NRm::TTxState> tx
, TIntrusivePtr<NRm::TTaskState> task
, ui64 limit
, ui64 reasonableSpillingTreshold)
, ui64 limit)
: NYql::NDq::TGuaranteeQuotaManager(limit, limit)
, ResourceManager(std::move(resourceManager))
, MemoryPool(memoryPool)
, State(std::move(state))
, Tx(std::move(tx))
, Task(std::move(task))
, ReasonableSpillingTreshold(reasonableSpillingTreshold)
{
}

Expand Down Expand Up @@ -57,7 +55,7 @@ struct TMemoryQuotaManager : public NYql::NDq::TGuaranteeQuotaManager {
}

bool IsReasonableToUseSpilling() const override {
return Tx->GetExtraMemoryAllocatedSize() >= ReasonableSpillingTreshold;
return Task->IsReasonableToStartSpilling();
}

TString MemoryConsumptionDetails() const override {
Expand Down Expand Up @@ -88,7 +86,6 @@ class TKqpCaFactory : public IKqpNodeComputeActorFactory {
std::atomic<ui64> MkqlLightProgramMemoryLimit = 0;
std::atomic<ui64> MkqlHeavyProgramMemoryLimit = 0;
std::atomic<ui64> MinChannelBufferSize = 0;
std::atomic<ui64> ReasonableSpillingTreshold = 0;

public:
TKqpCaFactory(const NKikimrConfig::TTableServiceConfig::TResourceManager& config,
Expand All @@ -107,7 +104,6 @@ class TKqpCaFactory : public IKqpNodeComputeActorFactory {
MkqlLightProgramMemoryLimit.store(config.GetMkqlLightProgramMemoryLimit());
MkqlHeavyProgramMemoryLimit.store(config.GetMkqlHeavyProgramMemoryLimit());
MinChannelBufferSize.store(config.GetMinChannelBufferSize());
ReasonableSpillingTreshold.store(config.GetReasonableSpillingTreshold());
}

TActorStartResult CreateKqpComputeActor(TCreateArgs&& args) override {
Expand Down Expand Up @@ -158,8 +154,7 @@ class TKqpCaFactory : public IKqpNodeComputeActorFactory {
std::move(args.State),
std::move(args.TxInfo),
std::move(task),
limit,
ReasonableSpillingTreshold.load());
limit);

auto runtimeSettings = args.RuntimeSettings;
runtimeSettings.ExtraMemoryAllocationPool = args.MemoryPool;
Expand Down
1 change: 1 addition & 0 deletions ydb/core/kqp/compute_actor/kqp_compute_actor_factory.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -122,6 +122,7 @@ struct IKqpNodeComputeActorFactory {
const TInstant& Deadline;
const bool ShareMailbox;
const TMaybe<NYql::NDqProto::TRlPath>& RlPath;

TComputeStagesWithScan* ComputesByStages = nullptr;
std::shared_ptr<IKqpNodeState> State = nullptr;
TComputeActorSchedulingOptions SchedulingOptions = {};
Expand Down
6 changes: 4 additions & 2 deletions ydb/core/kqp/executer_actor/kqp_planner.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -223,6 +223,7 @@ std::unique_ptr<TEvKqpNode::TEvStartKqpTasksRequest> TKqpPlanner::SerializeReque
}

request.SetSchedulerGroup(UserRequestContext->PoolId);
request.SetMemoryPoolPercent(UserRequestContext->PoolConfig->QueryMemoryLimitPercentPerNode);
Copy link
Collaborator

@GrigoriyPA GrigoriyPA Aug 7, 2024

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

В этом месте может быть UserRequestContext->PoolConfig = std::nullopt, лучше добавить на это проверку


return result;
}
Expand Down Expand Up @@ -351,7 +352,8 @@ TString TKqpPlanner::ExecuteDataComputeTask(ui64 taskId, ui32 computeTasksSize)
NYql::NDq::TComputeRuntimeSettings settings;
if (!TxInfo) {
TxInfo = MakeIntrusive<NRm::TTxState>(
TxId, TInstant::Now(), ResourceManager_->GetCounters());
TxId, TInstant::Now(), ResourceManager_->GetCounters(),
UserRequestContext->PoolId, UserRequestContext->PoolConfig->QueryMemoryLimitPercentPerNode);
Copy link
Collaborator

@GrigoriyPA GrigoriyPA Aug 7, 2024

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Тут тоже стоит проверить UserRequestContext->PoolConfig = std::nullopt

}

auto startResult = CaFactory_->CreateKqpComputeActor({
Expand All @@ -370,7 +372,7 @@ TString TKqpPlanner::ExecuteDataComputeTask(ui64 taskId, ui32 computeTasksSize)
.StatsMode = GetDqStatsMode(StatsMode),
.Deadline = Deadline,
.ShareMailbox = (computeTasksSize <= 1),
.RlPath = Nothing()
.RlPath = Nothing(),
});

if (const auto* rmResult = std::get_if<NRm::TKqpRMAllocateResult>(&startResult)) {
Expand Down
5 changes: 3 additions & 2 deletions ydb/core/kqp/node_service/kqp_node_service.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -205,7 +205,8 @@ class TKqpNodeService : public TActorBootstrapped<TKqpNodeService> {
}

TIntrusivePtr<NRm::TTxState> txInfo = MakeIntrusive<NRm::TTxState>(
txId, TInstant::Now(), ResourceManager_->GetCounters());
txId, TInstant::Now(), ResourceManager_->GetCounters(),
msg.GetSchedulerGroup(), msg.GetMemoryPoolPercent());

const ui32 tasksCount = msg.GetTasks().size();
for (auto& dqTask: *msg.MutableTasks()) {
Expand Down Expand Up @@ -246,7 +247,7 @@ class TKqpNodeService : public TActorBootstrapped<TKqpNodeService> {
.RlPath = rlPath,
.ComputesByStages = &computesByStage,
.State = State_,
.SchedulingOptions = std::move(schedulingOptions)
.SchedulingOptions = std::move(schedulingOptions),
});

if (const auto* rmResult = std::get_if<NRm::TKqpRMAllocateResult>(&result)) {
Expand Down
Loading