blob: cc0d7e7dfd520d6a59ab7bb2e9d359456c0ce741 [file] [log] [blame]
/*
* Licensed to the Apache Software Foundation (ASF) under one
* or more contributor license agreements. See the NOTICE file
* distributed with this work for additional information
* regarding copyright ownership. The ASF licenses this file
* to you under the Apache License, Version 2.0 (the
* "License"); you may not use this file except in compliance
* with the License. You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing,
* software distributed under the License is distributed on an
* "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
* KIND, either express or implied. See the License for the
* specific language governing permissions and limitations
* under the License.
*/
package org.apache.iotdb.db.mpp.plan.planner;
import org.apache.iotdb.db.conf.IoTDBDescriptor;
import org.apache.iotdb.db.engine.storagegroup.DataRegion;
import org.apache.iotdb.db.metadata.schemaregion.ISchemaRegion;
import org.apache.iotdb.db.mpp.exception.MemoryNotEnoughException;
import org.apache.iotdb.db.mpp.execution.driver.DataDriver;
import org.apache.iotdb.db.mpp.execution.driver.DataDriverContext;
import org.apache.iotdb.db.mpp.execution.driver.SchemaDriver;
import org.apache.iotdb.db.mpp.execution.driver.SchemaDriverContext;
import org.apache.iotdb.db.mpp.execution.fragment.FragmentInstanceContext;
import org.apache.iotdb.db.mpp.execution.fragment.FragmentInstanceStateMachine;
import org.apache.iotdb.db.mpp.execution.operator.Operator;
import org.apache.iotdb.db.mpp.execution.timer.ITimeSliceAllocator;
import org.apache.iotdb.db.mpp.plan.analyze.TypeProvider;
import org.apache.iotdb.db.mpp.plan.planner.plan.node.PlanNode;
import org.apache.iotdb.db.utils.SetThreadName;
import org.apache.iotdb.rpc.TSStatusCode;
import org.apache.iotdb.tsfile.read.filter.basic.Filter;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
/**
* Used to plan a fragment instance. Currently, we simply change it from PlanNode to executable
* Operator tree, but in the future, we may split one fragment instance into multiple pipeline to
* run a fragment instance parallel and take full advantage of multi-cores
*/
public class LocalExecutionPlanner {
private static final Logger LOGGER = LoggerFactory.getLogger(LocalExecutionPlanner.class);
/** allocated memory for operator execution */
private long freeMemoryForOperators =
IoTDBDescriptor.getInstance().getConfig().getAllocateMemoryForOperators();
public static LocalExecutionPlanner getInstance() {
return InstanceHolder.INSTANCE;
}
public DataDriver plan(
PlanNode plan,
TypeProvider types,
FragmentInstanceContext instanceContext,
Filter timeFilter,
DataRegion dataRegion)
throws MemoryNotEnoughException {
LocalExecutionPlanContext context = new LocalExecutionPlanContext(types, instanceContext);
Operator root = plan.accept(new OperatorTreeGenerator(), context);
// check whether current free memory is enough to execute current query
checkMemory(root, instanceContext.getStateMachine());
ITimeSliceAllocator timeSliceAllocator = context.getTimeSliceAllocator();
instanceContext
.getOperatorContexts()
.forEach(
operatorContext ->
operatorContext.setMaxRunTime(timeSliceAllocator.getMaxRunTime(operatorContext)));
DataDriverContext dataDriverContext =
new DataDriverContext(
instanceContext,
context.getPaths(),
timeFilter,
dataRegion,
context.getSourceOperators());
instanceContext.setDriverContext(dataDriverContext);
return new DataDriver(root, context.getSinkHandle(), dataDriverContext);
}
public SchemaDriver plan(
PlanNode plan, FragmentInstanceContext instanceContext, ISchemaRegion schemaRegion)
throws MemoryNotEnoughException {
SchemaDriverContext schemaDriverContext =
new SchemaDriverContext(instanceContext, schemaRegion);
instanceContext.setDriverContext(schemaDriverContext);
LocalExecutionPlanContext context = new LocalExecutionPlanContext(instanceContext);
Operator root = plan.accept(new OperatorTreeGenerator(), context);
// check whether current free memory is enough to execute current query
checkMemory(root, instanceContext.getStateMachine());
ITimeSliceAllocator timeSliceAllocator = context.getTimeSliceAllocator();
instanceContext
.getOperatorContexts()
.forEach(
operatorContext ->
operatorContext.setMaxRunTime(timeSliceAllocator.getMaxRunTime(operatorContext)));
return new SchemaDriver(root, context.getSinkHandle(), schemaDriverContext);
}
private void checkMemory(Operator root, FragmentInstanceStateMachine stateMachine)
throws MemoryNotEnoughException {
// if it is disabled, just return
if (!IoTDBDescriptor.getInstance().getConfig().isEnableQueryMemoryEstimation()) {
return;
}
long estimatedMemorySize = root.calculateMaxPeekMemory();
synchronized (this) {
if (estimatedMemorySize > freeMemoryForOperators) {
throw new MemoryNotEnoughException(
String.format(
"There is not enough memory to execute current fragment instance, current remaining free memory is %d, estimated memory usage for current fragment instance is %d",
freeMemoryForOperators, estimatedMemorySize),
TSStatusCode.MEMORY_NOT_ENOUGH.getStatusCode());
} else {
freeMemoryForOperators -= estimatedMemorySize;
LOGGER.info(
String.format(
"[ConsumeMemory] consume: %d, current remaining memory: %d",
estimatedMemorySize, freeMemoryForOperators));
}
}
stateMachine.addStateChangeListener(
newState -> {
if (newState.isDone()) {
try (SetThreadName fragmentInstanceName =
new SetThreadName(stateMachine.getFragmentInstanceId().getFullId())) {
synchronized (this) {
this.freeMemoryForOperators += estimatedMemorySize;
LOGGER.info(
String.format(
"[ReleaseMemory] release: %d, current remaining memory: %d",
estimatedMemorySize, freeMemoryForOperators));
}
}
}
});
}
private static class InstanceHolder {
private InstanceHolder() {}
private static final LocalExecutionPlanner INSTANCE = new LocalExecutionPlanner();
}
}