[
https://issues.apache.org/jira/browse/DRILL-5956?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17059212#comment-17059212
]
ASF GitHub Bot commented on DRILL-5956:
---------------------------------------
paul-rogers commented on pull request #1888: DRILL-5956: Add Storage Plugin for
Apache Druid
URL: https://github.com/apache/drill/pull/1888#discussion_r352262941
##########
File path:
contrib/storage-druid/src/main/java/org/apache/drill/exec/store/druid/DruidGroupScan.java
##########
@@ -0,0 +1,247 @@
+/*
+ * 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.drill.exec.store.druid;
+
+import com.fasterxml.jackson.annotation.JacksonInject;
+import com.fasterxml.jackson.annotation.JsonCreator;
+import com.fasterxml.jackson.annotation.JsonProperty;
+import com.fasterxml.jackson.annotation.JsonTypeName;
+import com.fasterxml.jackson.annotation.JsonIgnore;
+
+import org.apache.drill.common.exceptions.ExecutionSetupException;
+import org.apache.drill.common.expression.SchemaPath;
+import org.apache.drill.exec.physical.EndpointAffinity;
+import org.apache.drill.exec.physical.base.AbstractGroupScan;
+import org.apache.drill.exec.physical.base.GroupScan;
+import org.apache.drill.exec.physical.base.PhysicalOperator;
+import org.apache.drill.exec.physical.base.ScanStats;
+import org.apache.drill.exec.proto.CoordinationProtos;
+import org.apache.drill.exec.store.StoragePluginRegistry;
+
+import org.apache.drill.exec.store.schedule.AffinityCreator;
+import org.apache.drill.exec.store.schedule.AssignmentCreator;
+import org.apache.drill.exec.store.schedule.CompleteWork;
+import org.apache.drill.exec.store.schedule.EndpointByteMap;
+import org.apache.drill.exec.store.schedule.EndpointByteMapImpl;
+import org.apache.drill.shaded.guava.com.google.common.base.Preconditions;
+import org.apache.drill.shaded.guava.com.google.common.collect.ListMultimap;
+import org.apache.drill.shaded.guava.com.google.common.collect.Lists;
+
+import java.io.IOException;
+import java.util.ArrayList;
+import java.util.List;
+import java.util.UUID;
+
+@JsonTypeName("druid-scan")
+public class DruidGroupScan extends AbstractGroupScan {
+
+ private static final org.slf4j.Logger logger =
org.slf4j.LoggerFactory.getLogger(DruidGroupScan.class);
+ private static final long DEFAULT_TABLET_SIZE = 1000;
+
+ private DruidStoragePluginConfig storagePluginConfig;
+ private List<SchemaPath> columns;
+ private DruidScanSpec scanSpec;
+ private DruidStoragePlugin storagePlugin;
+ private boolean filterPushedDown = false;
+ private List<DruidWork> druidWorkList = new ArrayList<>();
+ private ListMultimap<Integer,DruidWork> assignments;
+ private List<EndpointAffinity> affinities;
+ private String objectName;
+
+ @JsonCreator
+ public DruidGroupScan(@JsonProperty("scanSpec") DruidScanSpec scanSpec,
+ @JsonProperty("storagePluginConfig")
DruidStoragePluginConfig storagePluginConfig,
+ @JsonProperty("columns") List<SchemaPath> columns,
+ @JacksonInject StoragePluginRegistry pluginRegistry)
+ throws IOException, ExecutionSetupException {
+ this((DruidStoragePlugin)
pluginRegistry.getPlugin(storagePluginConfig), scanSpec, columns);
+ int columnSize = (columns == null) ? 0 : columns.size();
+ }
+
+ public DruidGroupScan(DruidStoragePlugin storagePlugin, DruidScanSpec
scanSpec,
+ List<SchemaPath> columns) {
+ super("someuser");
+ objectName = UUID.randomUUID().toString();
+ this.storagePlugin = storagePlugin;
+ this.storagePluginConfig = storagePlugin.getConfig();
+ this.scanSpec = scanSpec;
+ this.columns = columns == null || columns.size() == 0? ALL_COLUMNS :
columns;
+ init();
+ }
+
+ /**
+ * Private constructor, used for cloning.
+ * @param that The DruidGroupScan to clone
+ */
+ private DruidGroupScan(DruidGroupScan that) {
+ super(that);
+ this.columns = that.columns;
+ this.scanSpec = that.scanSpec;
+ this.storagePlugin = that.storagePlugin;
+ this.storagePluginConfig = that.storagePluginConfig;
+ this.filterPushedDown = that.filterPushedDown;
+ this.druidWorkList = that.druidWorkList;
+ this.assignments = that.assignments;
+ this.objectName = that.objectName;
+ }
+
+ @Override
+ public GroupScan clone(List<SchemaPath> columns) {
+ DruidGroupScan newScan = new DruidGroupScan(this);
+ newScan.columns = columns;
+ return newScan;
+ }
+
+ @Override
+ public List<EndpointAffinity> getOperatorAffinity() {
+ if (affinities == null) {
+ affinities = AffinityCreator.getAffinityMap(druidWorkList);
+ }
+ return affinities;
+ }
+
+ @Override
+ public boolean canPushdownProjects(List<SchemaPath> columns) {
+ return true;
+ }
+
+ @JsonIgnore
+ public boolean isFilterPushedDown() {
+ return filterPushedDown;
+ }
+
+ @JsonIgnore
+ public void setFilterPushedDown(boolean filterPushedDown) {
+ this.filterPushedDown = filterPushedDown;
+ }
+
+ private void init() {
+ logger.debug("Adding Druid Work for Table - " + getTableName() + "
Filter - " + getScanSpec().getFilters());
+
+ DruidWork druidWork =
+ new DruidWork(
+ new DruidSubScan.DruidSubScanSpec(
+ getTableName(),
+ getScanSpec().getFilters()
+ )
+ );
+ druidWorkList.add(druidWork);
+ }
+
+ private static class DruidWork implements CompleteWork {
+
+ private EndpointByteMapImpl byteMap = new EndpointByteMapImpl();
+
+ private DruidSubScan.DruidSubScanSpec druidSubScanSpec;
+
+ public DruidWork(DruidSubScan.DruidSubScanSpec druidSubScanSpec) {
+ this.druidSubScanSpec = druidSubScanSpec;
+ }
+
+ public DruidSubScan.DruidSubScanSpec getDruidSubScanSpec() {
+ return druidSubScanSpec;
+ }
+
+ @Override
+ public long getTotalBytes() {
+ return DEFAULT_TABLET_SIZE;
+ }
+
+ @Override
+ public EndpointByteMap getByteMap() {
+ return byteMap;
+ }
+
+ @Override
+ public int compareTo(CompleteWork o) {
+ return 0;
+ }
+ }
+
+ //TODO - MAY GET MORE PRECISE COUNT FROM DRUID ITSELF.
+ public ScanStats getScanStats() {
Review comment:
Good that we have a large record count. Multiply that number by some guessed
average row width (say 100 or 200) to get a byte count.
Then, be sure to adjust the estimate as push-downs are applied. Pushing down
columns should reduce the average row width (lowering cost) so that Calcite
prefers it. Filter push-down should lower row count (again, so Calcite prefers
it.)
(Sorry this is so complex; we should have a simpler API.)
----------------------------------------------------------------
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.
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
> Add Storage Plugin for Apache Druid
> -----------------------------------
>
> Key: DRILL-5956
> URL: https://issues.apache.org/jira/browse/DRILL-5956
> Project: Apache Drill
> Issue Type: Wish
> Components: Storage - Other
> Reporter: Jiaqi Liu
> Priority: Major
> Labels: Enhancement, Storage-Plugin
> Fix For: 1.18.0
>
>
> As more and more companies are using Druid for mission-critical industrial
> products, Drill could gain much more popularity with Druid as one of its
> supported storage plugin so that uses could easily bind Druid cluster to
> running Drill instance
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
