albertobastos commented on code in PR #15694: URL: https://github.com/apache/pinot/pull/15694#discussion_r2073031413
########## pinot-integration-tests/src/test/java/org/apache/pinot/integration/tests/MultiStageEngineSmallBufferTest.java: ########## @@ -0,0 +1,181 @@ +/** + * 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.pinot.integration.tests; + +import com.fasterxml.jackson.databind.JsonNode; +import java.io.File; +import java.util.ArrayList; +import java.util.Collections; +import java.util.List; +import java.util.concurrent.CountDownLatch; +import org.apache.commons.io.FileUtils; +import org.apache.helix.model.HelixConfigScope; +import org.apache.helix.model.builder.HelixConfigScopeBuilder; +import org.apache.pinot.spi.config.table.TableConfig; +import org.apache.pinot.spi.data.Schema; +import org.apache.pinot.spi.env.PinotConfiguration; +import org.apache.pinot.spi.utils.CommonConstants; +import org.apache.pinot.util.TestUtils; +import org.testng.annotations.AfterClass; +import org.testng.annotations.BeforeClass; +import org.testng.annotations.BeforeMethod; +import org.testng.annotations.Test; + +import static org.apache.pinot.spi.utils.CommonConstants.MultiStageQueryRunner.KEY_OF_ENABLE_DATA_BLOCK_PAYLOAD_SPLIT; +import static org.apache.pinot.spi.utils.CommonConstants.MultiStageQueryRunner.KEY_OF_MAX_INBOUND_QUERY_DATA_BLOCK_SIZE_BYTES; +import static org.junit.Assert.assertEquals; +import static org.junit.Assert.assertNotNull; + + +public class MultiStageEngineSmallBufferTest extends BaseClusterIntegrationTestSet { + + private static final String SCHEMA_FILE_NAME = "On_Time_On_Time_Performance_2014_100k_subset_nonulls.schema"; + private static final int NUM_SERVERS = 4; + private static final int INBOUND_BLOCK_SIZE = 256; + + @BeforeClass + public void setUp() + throws Exception { + TestUtils.ensureDirectoriesExistAndEmpty(_tempDir, _segmentDir, _tarDir); + + // Start the Pinot cluster + startZk(); + startController(); + + // Set the multi-stage max server query threads for the cluster, so that we can test the query queueing logic + // in the MultiStageBrokerRequestHandler + HelixConfigScope scope = + new HelixConfigScopeBuilder(HelixConfigScope.ConfigScopeProperty.CLUSTER).forCluster(getHelixClusterName()) + .build(); + _helixManager.getConfigAccessor() + .set(scope, CommonConstants.Helix.CONFIG_OF_MULTI_STAGE_ENGINE_MAX_SERVER_QUERY_THREADS, "30"); + + startBroker(); + startServer(); + + // Create and upload the schema and table config + Schema schema = createSchema(); + addSchema(schema); + TableConfig tableConfig = createOfflineTableConfig(); + addTableConfig(tableConfig); + + // Unpack the Avro files + List<File> avroFiles = unpackAvroData(_tempDir); + + // Create and upload segments + ClusterIntegrationTestUtils.buildSegmentsFromAvro(avroFiles, tableConfig, schema, 0, _segmentDir, _tarDir); + uploadSegments(getTableName(), _tarDir); + + // Set up the H2 connection + setUpH2Connection(avroFiles); + + // Initialize the query generator + setUpQueryGenerator(avroFiles); + + // Wait for all documents loaded + waitForAllDocsLoaded(600_000L); + } + + @Override + protected void startServer() + throws Exception { + startServers(NUM_SERVERS); + } + + @AfterClass + public void tearDown() + throws Exception { + dropOfflineTable(DEFAULT_TABLE_NAME); + + stopServer(); + stopBroker(); + stopController(); + stopZk(); + + FileUtils.deleteDirectory(_tempDir); + } + + @Override + protected String getSchemaFileName() { + return SCHEMA_FILE_NAME; + } + + @BeforeMethod + @Override + public void resetMultiStage() { + setUseMultiStageQueryEngine(true); + } + + @Override + protected void overrideBrokerConf(PinotConfiguration brokerConf) { + brokerConf.setProperty(KEY_OF_MAX_INBOUND_QUERY_DATA_BLOCK_SIZE_BYTES, INBOUND_BLOCK_SIZE); + brokerConf.setProperty(KEY_OF_ENABLE_DATA_BLOCK_PAYLOAD_SPLIT, true); + } + + @Override + protected void overrideServerConf(PinotConfiguration serverConf) { + serverConf.setProperty(KEY_OF_MAX_INBOUND_QUERY_DATA_BLOCK_SIZE_BYTES, INBOUND_BLOCK_SIZE); + serverConf.setProperty(KEY_OF_ENABLE_DATA_BLOCK_PAYLOAD_SPLIT, true); + } + + @Test + public void testConcurrentSplittedMailboxes() + throws Exception { + int numClients = 32; + + String query = + "SELECT ActualElapsedTime FROM mytable ORDER BY ActualElapsedTime DESC LIMIT 100"; + String expected = "[[678],[668],[662],[658],[651],[650],[647],[629],[625],[625],[621],[617],[610],[610],[607],[605]" + + ",[603],[582],[578],[576],[574],[572],[572],[566],[565],[564],[558],[555],[555],[554],[554],[553],[552],[550]" + + ",[549],[544],[543],[541],[540],[538],[537],[535],[533],[532],[526],[521],[520],[519],[518],[516],[515],[514]" + + ",[508],[508],[507],[505],[505],[502],[502],[499],[499],[498],[495],[494],[494],[493],[490],[487],[487],[484]" + + ",[484],[481],[481],[480],[479],[478],[478],[477],[473],[472],[471],[468],[465],[464],[464],[463],[461],[461]" + + ",[460],[459],[459],[458],[457],[456],[453],[453],[451],[451],[450],[450]]"; + + List<String> results = Collections.synchronizedList(new ArrayList<>()); + + CountDownLatch startLatch = new CountDownLatch(1); + CountDownLatch doneLatch = new CountDownLatch(numClients); + + for (int i = 0; i < numClients; i++) { + new Thread(() -> { Review Comment: You're right. Moved to an ExecutorService to shut them down afterwards. ########## pinot-integration-tests/src/test/java/org/apache/pinot/integration/tests/MultiStageEngineSmallBufferTest.java: ########## @@ -0,0 +1,181 @@ +/** + * 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.pinot.integration.tests; + +import com.fasterxml.jackson.databind.JsonNode; +import java.io.File; +import java.util.ArrayList; +import java.util.Collections; +import java.util.List; +import java.util.concurrent.CountDownLatch; +import org.apache.commons.io.FileUtils; +import org.apache.helix.model.HelixConfigScope; +import org.apache.helix.model.builder.HelixConfigScopeBuilder; +import org.apache.pinot.spi.config.table.TableConfig; +import org.apache.pinot.spi.data.Schema; +import org.apache.pinot.spi.env.PinotConfiguration; +import org.apache.pinot.spi.utils.CommonConstants; +import org.apache.pinot.util.TestUtils; +import org.testng.annotations.AfterClass; +import org.testng.annotations.BeforeClass; +import org.testng.annotations.BeforeMethod; +import org.testng.annotations.Test; + +import static org.apache.pinot.spi.utils.CommonConstants.MultiStageQueryRunner.KEY_OF_ENABLE_DATA_BLOCK_PAYLOAD_SPLIT; +import static org.apache.pinot.spi.utils.CommonConstants.MultiStageQueryRunner.KEY_OF_MAX_INBOUND_QUERY_DATA_BLOCK_SIZE_BYTES; +import static org.junit.Assert.assertEquals; +import static org.junit.Assert.assertNotNull; + + +public class MultiStageEngineSmallBufferTest extends BaseClusterIntegrationTestSet { + + private static final String SCHEMA_FILE_NAME = "On_Time_On_Time_Performance_2014_100k_subset_nonulls.schema"; + private static final int NUM_SERVERS = 4; + private static final int INBOUND_BLOCK_SIZE = 256; + + @BeforeClass + public void setUp() + throws Exception { + TestUtils.ensureDirectoriesExistAndEmpty(_tempDir, _segmentDir, _tarDir); + + // Start the Pinot cluster + startZk(); + startController(); + + // Set the multi-stage max server query threads for the cluster, so that we can test the query queueing logic + // in the MultiStageBrokerRequestHandler + HelixConfigScope scope = + new HelixConfigScopeBuilder(HelixConfigScope.ConfigScopeProperty.CLUSTER).forCluster(getHelixClusterName()) + .build(); + _helixManager.getConfigAccessor() + .set(scope, CommonConstants.Helix.CONFIG_OF_MULTI_STAGE_ENGINE_MAX_SERVER_QUERY_THREADS, "30"); + + startBroker(); + startServer(); + + // Create and upload the schema and table config + Schema schema = createSchema(); + addSchema(schema); + TableConfig tableConfig = createOfflineTableConfig(); + addTableConfig(tableConfig); + + // Unpack the Avro files + List<File> avroFiles = unpackAvroData(_tempDir); + + // Create and upload segments + ClusterIntegrationTestUtils.buildSegmentsFromAvro(avroFiles, tableConfig, schema, 0, _segmentDir, _tarDir); + uploadSegments(getTableName(), _tarDir); + + // Set up the H2 connection + setUpH2Connection(avroFiles); + + // Initialize the query generator + setUpQueryGenerator(avroFiles); + + // Wait for all documents loaded + waitForAllDocsLoaded(600_000L); + } + + @Override + protected void startServer() + throws Exception { + startServers(NUM_SERVERS); + } + + @AfterClass + public void tearDown() + throws Exception { + dropOfflineTable(DEFAULT_TABLE_NAME); + + stopServer(); + stopBroker(); + stopController(); + stopZk(); + + FileUtils.deleteDirectory(_tempDir); + } + + @Override + protected String getSchemaFileName() { + return SCHEMA_FILE_NAME; + } + + @BeforeMethod + @Override + public void resetMultiStage() { + setUseMultiStageQueryEngine(true); + } + + @Override + protected void overrideBrokerConf(PinotConfiguration brokerConf) { + brokerConf.setProperty(KEY_OF_MAX_INBOUND_QUERY_DATA_BLOCK_SIZE_BYTES, INBOUND_BLOCK_SIZE); + brokerConf.setProperty(KEY_OF_ENABLE_DATA_BLOCK_PAYLOAD_SPLIT, true); + } + + @Override + protected void overrideServerConf(PinotConfiguration serverConf) { + serverConf.setProperty(KEY_OF_MAX_INBOUND_QUERY_DATA_BLOCK_SIZE_BYTES, INBOUND_BLOCK_SIZE); + serverConf.setProperty(KEY_OF_ENABLE_DATA_BLOCK_PAYLOAD_SPLIT, true); + } + + @Test + public void testConcurrentSplittedMailboxes() + throws Exception { + int numClients = 32; + + String query = + "SELECT ActualElapsedTime FROM mytable ORDER BY ActualElapsedTime DESC LIMIT 100"; + String expected = "[[678],[668],[662],[658],[651],[650],[647],[629],[625],[625],[621],[617],[610],[610],[607],[605]" + + ",[603],[582],[578],[576],[574],[572],[572],[566],[565],[564],[558],[555],[555],[554],[554],[553],[552],[550]" + + ",[549],[544],[543],[541],[540],[538],[537],[535],[533],[532],[526],[521],[520],[519],[518],[516],[515],[514]" + + ",[508],[508],[507],[505],[505],[502],[502],[499],[499],[498],[495],[494],[494],[493],[490],[487],[487],[484]" + + ",[484],[481],[481],[480],[479],[478],[478],[477],[473],[472],[471],[468],[465],[464],[464],[463],[461],[461]" + + ",[460],[459],[459],[458],[457],[456],[453],[453],[451],[451],[450],[450]]"; + + List<String> results = Collections.synchronizedList(new ArrayList<>()); + + CountDownLatch startLatch = new CountDownLatch(1); + CountDownLatch doneLatch = new CountDownLatch(numClients); + + for (int i = 0; i < numClients; i++) { + new Thread(() -> { + try { + startLatch.await(); + JsonNode jsonNode = postQuery(query); + assertNotNull(jsonNode); + String actual = jsonNode.get("resultTable").get("rows").toString(); + results.add(actual); Review Comment: Yeah, that's a legit concern, but certainly a hard one to overcome. For now I just added your suggestion and changed the test to run the requests-and-check N times, with N set as 50 (one iteration took locally about 15+1 seconds, 50 iterations took about 18+6 which I find still acceptable). ########## pinot-query-runtime/src/main/java/org/apache/pinot/query/mailbox/GrpcSendingMailbox.java: ########## @@ -67,28 +66,22 @@ public class GrpcSendingMailbox implements SendingMailbox { private final long _deadlineMs; private final StatMap<MailboxSendOperator.StatKey> _statMap; private final MailboxStatusObserver _statusObserver = new MailboxStatusObserver(); - private final boolean _splitBlocks; private final int _maxByteStringSize; + private final Sender _sender; private StreamObserver<MailboxContent> _contentObserver; public GrpcSendingMailbox( PinotConfiguration config, String id, ChannelManager channelManager, String hostname, int port, long deadlineMs, - StatMap<MailboxSendOperator.StatKey> statMap) { + StatMap<MailboxSendOperator.StatKey> statMap, boolean splitBlocks, int maxByteStringSize) { _id = id; _channelManager = channelManager; _hostname = hostname; _port = port; _deadlineMs = deadlineMs; _statMap = statMap; - _splitBlocks = config.getProperty( - CommonConstants.MultiStageQueryRunner.KEY_OF_ENABLE_DATA_BLOCK_PAYLOAD_SPLIT, - CommonConstants.MultiStageQueryRunner.DEFAULT_ENABLE_DATA_BLOCK_PAYLOAD_SPLIT); - // so far we ensure payload is not bigger than maxBlockSize/2, we can fine tune this later - _maxByteStringSize = Math.max(config.getProperty( - CommonConstants.MultiStageQueryRunner.KEY_OF_MAX_INBOUND_QUERY_DATA_BLOCK_SIZE_BYTES, - CommonConstants.MultiStageQueryRunner.DEFAULT_MAX_INBOUND_QUERY_DATA_BLOCK_SIZE_BYTES - ) / 2, 1); + _maxByteStringSize = maxByteStringSize; Review Comment: You're right, moved it to a private member in the sender that requires it. ########## pinot-query-runtime/src/main/java/org/apache/pinot/query/mailbox/GrpcSendingMailbox.java: ########## @@ -67,28 +66,22 @@ public class GrpcSendingMailbox implements SendingMailbox { private final long _deadlineMs; private final StatMap<MailboxSendOperator.StatKey> _statMap; private final MailboxStatusObserver _statusObserver = new MailboxStatusObserver(); - private final boolean _splitBlocks; private final int _maxByteStringSize; + private final Sender _sender; private StreamObserver<MailboxContent> _contentObserver; public GrpcSendingMailbox( PinotConfiguration config, String id, ChannelManager channelManager, String hostname, int port, long deadlineMs, - StatMap<MailboxSendOperator.StatKey> statMap) { + StatMap<MailboxSendOperator.StatKey> statMap, boolean splitBlocks, int maxByteStringSize) { Review Comment: Yup, just used `splitBlocks` locally and assigning `maxByteStringSize=0` if we are not splitting them. ########## pinot-query-runtime/src/main/java/org/apache/pinot/query/mailbox/GrpcSendingMailbox.java: ########## @@ -319,4 +279,78 @@ static List<ByteString> toByteStrings(List<ByteBuffer> bytes, int maxByteStringS return result; } + + private static abstract class Sender { + protected abstract void send(MseBlock block, List<DataBuffer> serializedStats) + throws IOException; + } + + private static class SplitSender extends Sender { + private final GrpcSendingMailbox _mailbox; + + public SplitSender(GrpcSendingMailbox mailbox) { + _mailbox = mailbox; + } + + @Override + protected void send(MseBlock block, List<DataBuffer> serializedStats) + throws IOException { + _mailbox._statMap.merge(MailboxSendOperator.StatKey.RAW_MESSAGES, 1); Review Comment: I thought about that but didn't find the way to reuse the code without altering the order in which log traces and metrics get updated. But if you also saw it, I guess it is worth the change for the sake of readability. I split the code into two methods: a `send(MseBlock)` implemented by the abstract class and a `processAndSend(DataBlock)` with the specific logic for Split/NonSplit implementations. With this approach, the `debug` trace and `SERIALIZED_BYTES` metrics are not update until the contents has been sent instead of between calculating the payload byte size and effectively send it. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: commits-unsubscr...@pinot.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org --------------------------------------------------------------------- To unsubscribe, e-mail: commits-unsubscr...@pinot.apache.org For additional commands, e-mail: commits-h...@pinot.apache.org
