You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@flink.apache.org by GitBox <gi...@apache.org> on 2022/08/04 04:56:20 UTC
[GitHub] [flink] SmirAlex opened a new pull request, #20447: Flink 28532
SmirAlex opened a new pull request, #20447:
URL: https://github.com/apache/flink/pull/20447
<!--
*Thank you very much for contributing to Apache Flink - we are happy that you want to help us improve Flink. To help the community review your contribution in the best possible way, please go through the checklist below, which will get the contribution into a shape in which it can be best reviewed.*
*Please understand that we do not do this to make contributions to Flink a hassle. In order to uphold a high standard of quality for code contributions, while at the same time managing a large number of contributions, we need contributors to prepare the contributions well, and give reviewers enough contextual information for the review. Please also understand that contributions that do not follow this guide will take longer to review and thus typically be picked up with lower priority by the community.*
## Contribution Checklist
- Make sure that the pull request corresponds to a [JIRA issue](https://issues.apache.org/jira/projects/FLINK/issues). Exceptions are made for typos in JavaDoc or documentation files, which need no JIRA issue.
- Name the pull request in the form "[FLINK-XXXX] [component] Title of the pull request", where *FLINK-XXXX* should be replaced by the actual issue number. Skip *component* if you are unsure about which is the best component.
Typo fixes that have no associated JIRA issue should be named following this pattern: `[hotfix] [docs] Fix typo in event time introduction` or `[hotfix] [javadocs] Expand JavaDoc for PuncuatedWatermarkGenerator`.
- Fill out the template below to describe the changes contributed by the pull request. That will give reviewers the context they need to do the review.
- Make sure that the change passes the automated tests, i.e., `mvn clean verify` passes. You can set up Azure Pipelines CI to do that following [this guide](https://cwiki.apache.org/confluence/display/FLINK/Azure+Pipelines#AzurePipelines-Tutorial:SettingupAzurePipelinesforaforkoftheFlinkrepository).
- Each pull request should address only one issue, not mix up code from multiple issues.
- Each commit in the pull request has a meaningful commit message (including the JIRA id)
- Once all items of the checklist are addressed, remove the above text and this checklist, leaving only the filled out template below.
**(The sections below can be removed for hotfixes of typos)**
-->
## What is the purpose of the change
This pull request introduces support of lookup join using FullCachingLookupProvider that uses InputFormats for data loading. Also it contains change of LookupJoinRunner, so now it uses projection to calculate cache keys + other commits with partial caching (they will gone after future rebasing).
## Brief change log
- Add `LookupFullCache` as an implementation of interface `LookupCache`, so the same logic with caching can be reused in runners.
- There are 2 copies of full caches (old and new) at the process of reloading, so no blocking is required, but on the other hand, it requires twice more memory.
- Reloading happens in another thread. If exception will be thrown from this thread, main thread will also stop.
- If there are more than one InputSplit from InputFormat, reloading will be performed in parallel. Total number of threads for reloading = min(numOfSplit, numOfCores).
## Verifying this change
This change added tests and can be verified as follows:
- `LookupJoinHarnessTest` - verifies correct result (the same as with NONE and PARTIAL caching)
- Will be more...
## Does this pull request potentially affect one of the following parts:
- Dependencies (does it add or upgrade a dependency): (no)
- The public API, i.e., is any changed class annotated with `@Public(Evolving)`: (no)
- The serializers: (yes)
- The runtime per-record code paths (performance sensitive): (yes)
- Anything that affects deployment or recovery: JobManager (and its components), Checkpointing, Kubernetes/Yarn, ZooKeeper: (no)
- The S3 file system connector: (no)
## Documentation
- Does this pull request introduce a new feature? (yes)
- If yes, how is the feature documented? (JavaDocs)
--
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: issues-unsubscribe@flink.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [flink] SmirAlex commented on a diff in pull request #20447: [FLINK-28532][table] Support full caching in lookup join runner using InputFormats as scan runtime provider
Posted by GitBox <gi...@apache.org>.
SmirAlex commented on code in PR #20447:
URL: https://github.com/apache/flink/pull/20447#discussion_r941504036
##########
flink-table/flink-table-planner/src/main/java/org/apache/flink/table/planner/plan/utils/LookupJoinUtil.java:
##########
@@ -208,6 +217,21 @@ public static UserDefinedFunction getLookupFunction(
return new CachingAsyncLookupFunction(
partialCachingLookupProvider.getCache(),
partialCachingLookupProvider.createAsyncLookupFunction());
+ } else if (provider instanceof FullCachingLookupProvider) {
+ FullCachingLookupProvider fullCachingLookupProvider =
+ (FullCachingLookupProvider) provider;
+ RowType tableSourceRowType =
+ FlinkTypeFactory.toLogicalRowType(temporalTable.getRowType());
+ LookupFullCache fullCache =
+ classLoader == null
Review Comment:
No, trace of past implementations
--
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: issues-unsubscribe@flink.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [flink] SmirAlex commented on a diff in pull request #20447: [FLINK-28532][table] Support full caching in lookup join runner using InputFormats as scan runtime provider
Posted by GitBox <gi...@apache.org>.
SmirAlex commented on code in PR #20447:
URL: https://github.com/apache/flink/pull/20447#discussion_r941503346
##########
flink-table/flink-table-planner/src/main/java/org/apache/flink/table/planner/plan/utils/LookupJoinUtil.java:
##########
@@ -208,6 +217,21 @@ public static UserDefinedFunction getLookupFunction(
return new CachingAsyncLookupFunction(
partialCachingLookupProvider.getCache(),
partialCachingLookupProvider.createAsyncLookupFunction());
+ } else if (provider instanceof FullCachingLookupProvider) {
Review Comment:
Yes, it is probably because of problems after rebase
--
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: issues-unsubscribe@flink.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [flink] PatrickRen merged pull request #20447: [FLINK-28532][table] Support full caching in lookup join runner using InputFormats as scan runtime provider
Posted by GitBox <gi...@apache.org>.
PatrickRen merged PR #20447:
URL: https://github.com/apache/flink/pull/20447
--
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: issues-unsubscribe@flink.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [flink] SmirAlex commented on a diff in pull request #20447: [FLINK-28532][table] Support full caching in lookup join runner using InputFormats as scan runtime provider
Posted by GitBox <gi...@apache.org>.
SmirAlex commented on code in PR #20447:
URL: https://github.com/apache/flink/pull/20447#discussion_r941497728
##########
flink-table/flink-table-runtime/src/main/java/org/apache/flink/table/runtime/functions/table/lookup/fullcache/inputformat/InputFormatCacheLoader.java:
##########
@@ -0,0 +1,159 @@
+/*
+ * 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.flink.table.runtime.functions.table.lookup.fullcache.inputformat;
+
+import org.apache.flink.api.common.io.InputFormat;
+import org.apache.flink.configuration.Configuration;
+import org.apache.flink.core.io.InputSplit;
+import org.apache.flink.table.data.RowData;
+import org.apache.flink.table.runtime.functions.table.lookup.fullcache.CacheLoader;
+import org.apache.flink.table.runtime.keyselector.GenericRowDataKeySelector;
+import org.apache.flink.table.runtime.typeutils.RowDataSerializer;
+import org.apache.flink.util.InstantiationUtil;
+import org.apache.flink.util.Preconditions;
+
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+
+import java.io.IOException;
+import java.util.ArrayList;
+import java.util.Arrays;
+import java.util.Collection;
+import java.util.List;
+import java.util.concurrent.ConcurrentHashMap;
+import java.util.concurrent.ExecutorService;
+import java.util.concurrent.Executors;
+import java.util.concurrent.Future;
+import java.util.stream.Collectors;
+
+/** {@link CacheLoader} that used {@link InputFormat} for loading data into the cache. */
+public class InputFormatCacheLoader extends CacheLoader {
+ private static final long serialVersionUID = 1L;
+ private static final Logger LOG = LoggerFactory.getLogger(InputFormatCacheLoader.class);
+
+ private final InputFormat<RowData, InputSplit> initialInputFormat;
+ private final GenericRowDataKeySelector keySelector;
+ private final RowDataSerializer cacheEntriesSerializer;
+
+ private transient volatile List<InputSplitCacheLoadTask> cacheLoadTasks;
+ private transient Configuration parameters;
+
+ private volatile boolean isStopped;
+
+ public InputFormatCacheLoader(
+ InputFormat<RowData, ?> initialInputFormat,
+ GenericRowDataKeySelector keySelector,
+ RowDataSerializer cacheEntriesSerializer) {
+ // noinspection unchecked
+ this.initialInputFormat = (InputFormat<RowData, InputSplit>) initialInputFormat;
+ this.keySelector = keySelector;
+ this.cacheEntriesSerializer = cacheEntriesSerializer;
+ }
+
+ @Override
+ public void open(Configuration parameters) throws Exception {
+ super.open(parameters);
+ this.parameters = parameters;
+ this.initialInputFormat.configure(parameters);
+ }
+
+ @Override
+ protected void reloadCache() throws Exception {
+ InputSplit[] inputSplits = createInputSplits();
+ int numSplits = inputSplits.length;
+ // load data into the another copy of cache
+ // notice: it requires twice more memory, but on the other hand we don't need any blocking
+ // cache has default initialCapacity and loadFactor, but overridden concurrencyLevel
+ ConcurrentHashMap<RowData, Collection<RowData>> newCache =
+ new ConcurrentHashMap<>(16, 0.75f, getConcurrencyLevel(numSplits));
+ this.cacheLoadTasks =
+ Arrays.stream(inputSplits)
+ .map(split -> createCacheLoadTask(split, newCache))
+ .collect(Collectors.toList());
+
+ // run first task and start numTasks - 1 threads to run remaining tasks
+ ExecutorService cacheLoadTaskService = null;
+ List<Future<?>> futures = null;
+ if (numSplits > 1) {
+ futures = new ArrayList<>();
+ int numThreads = getConcurrencyLevel(numSplits) - 1;
+ cacheLoadTaskService = Executors.newFixedThreadPool(numThreads);
+ for (int i = 1; i < numSplits; i++) {
+ Future<?> future = cacheLoadTaskService.submit(cacheLoadTasks.get(i));
+ futures.add(future);
+ }
+ }
+ cacheLoadTasks.get(0).run();
+ if (cacheLoadTaskService != null) {
+ for (Future<?> future : futures) {
+ future.get(); // if any exception occurs it will be thrown here
Review Comment:
We have simple futures here, not CompletableFutures.
--
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: issues-unsubscribe@flink.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [flink] SmirAlex commented on a diff in pull request #20447: [FLINK-28532][table] Support full caching in lookup join runner using InputFormats as scan runtime provider
Posted by GitBox <gi...@apache.org>.
SmirAlex commented on code in PR #20447:
URL: https://github.com/apache/flink/pull/20447#discussion_r941496189
##########
flink-table/flink-table-runtime/src/main/java/org/apache/flink/table/runtime/functions/table/lookup/fullcache/inputformat/InputFormatCacheLoader.java:
##########
@@ -0,0 +1,159 @@
+/*
+ * 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.flink.table.runtime.functions.table.lookup.fullcache.inputformat;
+
+import org.apache.flink.api.common.io.InputFormat;
+import org.apache.flink.configuration.Configuration;
+import org.apache.flink.core.io.InputSplit;
+import org.apache.flink.table.data.RowData;
+import org.apache.flink.table.runtime.functions.table.lookup.fullcache.CacheLoader;
+import org.apache.flink.table.runtime.keyselector.GenericRowDataKeySelector;
+import org.apache.flink.table.runtime.typeutils.RowDataSerializer;
+import org.apache.flink.util.InstantiationUtil;
+import org.apache.flink.util.Preconditions;
+
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+
+import java.io.IOException;
+import java.util.ArrayList;
+import java.util.Arrays;
+import java.util.Collection;
+import java.util.List;
+import java.util.concurrent.ConcurrentHashMap;
+import java.util.concurrent.ExecutorService;
+import java.util.concurrent.Executors;
+import java.util.concurrent.Future;
+import java.util.stream.Collectors;
+
+/** {@link CacheLoader} that used {@link InputFormat} for loading data into the cache. */
+public class InputFormatCacheLoader extends CacheLoader {
+ private static final long serialVersionUID = 1L;
+ private static final Logger LOG = LoggerFactory.getLogger(InputFormatCacheLoader.class);
+
+ private final InputFormat<RowData, InputSplit> initialInputFormat;
+ private final GenericRowDataKeySelector keySelector;
+ private final RowDataSerializer cacheEntriesSerializer;
+
+ private transient volatile List<InputSplitCacheLoadTask> cacheLoadTasks;
+ private transient Configuration parameters;
+
+ private volatile boolean isStopped;
+
+ public InputFormatCacheLoader(
+ InputFormat<RowData, ?> initialInputFormat,
+ GenericRowDataKeySelector keySelector,
+ RowDataSerializer cacheEntriesSerializer) {
+ // noinspection unchecked
+ this.initialInputFormat = (InputFormat<RowData, InputSplit>) initialInputFormat;
+ this.keySelector = keySelector;
+ this.cacheEntriesSerializer = cacheEntriesSerializer;
+ }
+
+ @Override
+ public void open(Configuration parameters) throws Exception {
+ super.open(parameters);
+ this.parameters = parameters;
+ this.initialInputFormat.configure(parameters);
+ }
+
+ @Override
+ protected void reloadCache() throws Exception {
+ InputSplit[] inputSplits = createInputSplits();
+ int numSplits = inputSplits.length;
+ // load data into the another copy of cache
+ // notice: it requires twice more memory, but on the other hand we don't need any blocking
+ // cache has default initialCapacity and loadFactor, but overridden concurrencyLevel
+ ConcurrentHashMap<RowData, Collection<RowData>> newCache =
+ new ConcurrentHashMap<>(16, 0.75f, getConcurrencyLevel(numSplits));
+ this.cacheLoadTasks =
+ Arrays.stream(inputSplits)
+ .map(split -> createCacheLoadTask(split, newCache))
+ .collect(Collectors.toList());
+
+ // run first task and start numTasks - 1 threads to run remaining tasks
+ ExecutorService cacheLoadTaskService = null;
+ List<Future<?>> futures = null;
+ if (numSplits > 1) {
+ futures = new ArrayList<>();
+ int numThreads = getConcurrencyLevel(numSplits) - 1;
+ cacheLoadTaskService = Executors.newFixedThreadPool(numThreads);
+ for (int i = 1; i < numSplits; i++) {
+ Future<?> future = cacheLoadTaskService.submit(cacheLoadTasks.get(i));
+ futures.add(future);
+ }
+ }
+ cacheLoadTasks.get(0).run();
Review Comment:
By doing this we don't create one more thread. Parent thread anyway waits for all tasks to complete, so he can execute one of tasks to not idle.
--
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: issues-unsubscribe@flink.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [flink] PatrickRen commented on a diff in pull request #20447: [FLINK-28532][table] Support full caching in lookup join runner using InputFormats as scan runtime provider
Posted by GitBox <gi...@apache.org>.
PatrickRen commented on code in PR #20447:
URL: https://github.com/apache/flink/pull/20447#discussion_r941394266
##########
flink-table/flink-table-runtime/src/main/java/org/apache/flink/table/runtime/functions/table/lookup/fullcache/LookupFullCache.java:
##########
@@ -0,0 +1,96 @@
+/*
+ * 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.flink.table.runtime.functions.table.lookup.fullcache;
+
+import org.apache.flink.configuration.Configuration;
+import org.apache.flink.metrics.groups.CacheMetricGroup;
+import org.apache.flink.table.connector.source.lookup.LookupOptions.LookupCacheType;
+import org.apache.flink.table.connector.source.lookup.cache.LookupCache;
+import org.apache.flink.table.connector.source.lookup.cache.trigger.CacheReloadTrigger;
+import org.apache.flink.table.data.RowData;
+import org.apache.flink.util.Preconditions;
+
+import java.util.Collection;
+import java.util.Collections;
+
+/** Internal implementation of {@link LookupCache} for {@link LookupCacheType#FULL}. */
+public class LookupFullCache implements LookupCache {
+ private static final long serialVersionUID = 1L;
+
+ private final CacheLoader cacheLoader;
+ private final CacheReloadTrigger reloadTrigger;
+
+ private volatile boolean metricsOpened;
+ private transient volatile ReloadTriggerContext reloadTriggerContext;
+ private transient volatile Throwable reloadFailCause;
+
+ public LookupFullCache(CacheLoader cacheLoader, CacheReloadTrigger reloadTrigger) {
+ this.cacheLoader = Preconditions.checkNotNull(cacheLoader);
+ this.reloadTrigger = Preconditions.checkNotNull(reloadTrigger);
+ }
+
+ @Override
+ public synchronized void open(CacheMetricGroup metricGroup) {
+ if (!metricsOpened) {
Review Comment:
I think we should allow the cache being opened for multiple times, as different subtasks will pass in different instance of metric groups.
Also this open is not necessary to use `synchronized` if we removes the `metricsOpened`
##########
flink-table/flink-table-planner/src/main/java/org/apache/flink/table/planner/plan/utils/LookupJoinUtil.java:
##########
@@ -208,6 +217,21 @@ public static UserDefinedFunction getLookupFunction(
return new CachingAsyncLookupFunction(
partialCachingLookupProvider.getCache(),
partialCachingLookupProvider.createAsyncLookupFunction());
+ } else if (provider instanceof FullCachingLookupProvider) {
Review Comment:
This should be under `LookupFunctionProvider` branch
##########
flink-table/flink-table-runtime/src/main/java/org/apache/flink/table/runtime/functions/table/lookup/fullcache/inputformat/InputFormatCacheLoader.java:
##########
@@ -0,0 +1,159 @@
+/*
+ * 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.flink.table.runtime.functions.table.lookup.fullcache.inputformat;
+
+import org.apache.flink.api.common.io.InputFormat;
+import org.apache.flink.configuration.Configuration;
+import org.apache.flink.core.io.InputSplit;
+import org.apache.flink.table.data.RowData;
+import org.apache.flink.table.runtime.functions.table.lookup.fullcache.CacheLoader;
+import org.apache.flink.table.runtime.keyselector.GenericRowDataKeySelector;
+import org.apache.flink.table.runtime.typeutils.RowDataSerializer;
+import org.apache.flink.util.InstantiationUtil;
+import org.apache.flink.util.Preconditions;
+
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+
+import java.io.IOException;
+import java.util.ArrayList;
+import java.util.Arrays;
+import java.util.Collection;
+import java.util.List;
+import java.util.concurrent.ConcurrentHashMap;
+import java.util.concurrent.ExecutorService;
+import java.util.concurrent.Executors;
+import java.util.concurrent.Future;
+import java.util.stream.Collectors;
+
+/** {@link CacheLoader} that used {@link InputFormat} for loading data into the cache. */
+public class InputFormatCacheLoader extends CacheLoader {
+ private static final long serialVersionUID = 1L;
+ private static final Logger LOG = LoggerFactory.getLogger(InputFormatCacheLoader.class);
+
+ private final InputFormat<RowData, InputSplit> initialInputFormat;
+ private final GenericRowDataKeySelector keySelector;
+ private final RowDataSerializer cacheEntriesSerializer;
+
+ private transient volatile List<InputSplitCacheLoadTask> cacheLoadTasks;
+ private transient Configuration parameters;
+
+ private volatile boolean isStopped;
+
+ public InputFormatCacheLoader(
+ InputFormat<RowData, ?> initialInputFormat,
+ GenericRowDataKeySelector keySelector,
+ RowDataSerializer cacheEntriesSerializer) {
+ // noinspection unchecked
+ this.initialInputFormat = (InputFormat<RowData, InputSplit>) initialInputFormat;
+ this.keySelector = keySelector;
+ this.cacheEntriesSerializer = cacheEntriesSerializer;
+ }
+
+ @Override
+ public void open(Configuration parameters) throws Exception {
+ super.open(parameters);
+ this.parameters = parameters;
+ this.initialInputFormat.configure(parameters);
+ }
+
+ @Override
+ protected void reloadCache() throws Exception {
+ InputSplit[] inputSplits = createInputSplits();
+ int numSplits = inputSplits.length;
+ // load data into the another copy of cache
+ // notice: it requires twice more memory, but on the other hand we don't need any blocking
+ // cache has default initialCapacity and loadFactor, but overridden concurrencyLevel
+ ConcurrentHashMap<RowData, Collection<RowData>> newCache =
+ new ConcurrentHashMap<>(16, 0.75f, getConcurrencyLevel(numSplits));
+ this.cacheLoadTasks =
+ Arrays.stream(inputSplits)
+ .map(split -> createCacheLoadTask(split, newCache))
+ .collect(Collectors.toList());
+
+ // run first task and start numTasks - 1 threads to run remaining tasks
+ ExecutorService cacheLoadTaskService = null;
+ List<Future<?>> futures = null;
+ if (numSplits > 1) {
+ futures = new ArrayList<>();
+ int numThreads = getConcurrencyLevel(numSplits) - 1;
+ cacheLoadTaskService = Executors.newFixedThreadPool(numThreads);
+ for (int i = 1; i < numSplits; i++) {
+ Future<?> future = cacheLoadTaskService.submit(cacheLoadTasks.get(i));
+ futures.add(future);
+ }
+ }
+ cacheLoadTasks.get(0).run();
+ if (cacheLoadTaskService != null) {
+ for (Future<?> future : futures) {
+ future.get(); // if any exception occurs it will be thrown here
Review Comment:
What about using `CompletableFuture#allOf` or `FutureUtils#completeAll`?
##########
flink-table/flink-table-runtime/src/main/java/org/apache/flink/table/runtime/functions/table/lookup/fullcache/inputformat/InputFormatCacheLoader.java:
##########
@@ -0,0 +1,159 @@
+/*
+ * 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.flink.table.runtime.functions.table.lookup.fullcache.inputformat;
+
+import org.apache.flink.api.common.io.InputFormat;
+import org.apache.flink.configuration.Configuration;
+import org.apache.flink.core.io.InputSplit;
+import org.apache.flink.table.data.RowData;
+import org.apache.flink.table.runtime.functions.table.lookup.fullcache.CacheLoader;
+import org.apache.flink.table.runtime.keyselector.GenericRowDataKeySelector;
+import org.apache.flink.table.runtime.typeutils.RowDataSerializer;
+import org.apache.flink.util.InstantiationUtil;
+import org.apache.flink.util.Preconditions;
+
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+
+import java.io.IOException;
+import java.util.ArrayList;
+import java.util.Arrays;
+import java.util.Collection;
+import java.util.List;
+import java.util.concurrent.ConcurrentHashMap;
+import java.util.concurrent.ExecutorService;
+import java.util.concurrent.Executors;
+import java.util.concurrent.Future;
+import java.util.stream.Collectors;
+
+/** {@link CacheLoader} that used {@link InputFormat} for loading data into the cache. */
+public class InputFormatCacheLoader extends CacheLoader {
+ private static final long serialVersionUID = 1L;
+ private static final Logger LOG = LoggerFactory.getLogger(InputFormatCacheLoader.class);
+
+ private final InputFormat<RowData, InputSplit> initialInputFormat;
+ private final GenericRowDataKeySelector keySelector;
+ private final RowDataSerializer cacheEntriesSerializer;
+
+ private transient volatile List<InputSplitCacheLoadTask> cacheLoadTasks;
+ private transient Configuration parameters;
+
+ private volatile boolean isStopped;
+
+ public InputFormatCacheLoader(
+ InputFormat<RowData, ?> initialInputFormat,
+ GenericRowDataKeySelector keySelector,
+ RowDataSerializer cacheEntriesSerializer) {
+ // noinspection unchecked
+ this.initialInputFormat = (InputFormat<RowData, InputSplit>) initialInputFormat;
+ this.keySelector = keySelector;
+ this.cacheEntriesSerializer = cacheEntriesSerializer;
+ }
+
+ @Override
+ public void open(Configuration parameters) throws Exception {
+ super.open(parameters);
+ this.parameters = parameters;
+ this.initialInputFormat.configure(parameters);
+ }
+
+ @Override
+ protected void reloadCache() throws Exception {
+ InputSplit[] inputSplits = createInputSplits();
+ int numSplits = inputSplits.length;
+ // load data into the another copy of cache
+ // notice: it requires twice more memory, but on the other hand we don't need any blocking
+ // cache has default initialCapacity and loadFactor, but overridden concurrencyLevel
+ ConcurrentHashMap<RowData, Collection<RowData>> newCache =
+ new ConcurrentHashMap<>(16, 0.75f, getConcurrencyLevel(numSplits));
+ this.cacheLoadTasks =
+ Arrays.stream(inputSplits)
+ .map(split -> createCacheLoadTask(split, newCache))
+ .collect(Collectors.toList());
+
+ // run first task and start numTasks - 1 threads to run remaining tasks
+ ExecutorService cacheLoadTaskService = null;
+ List<Future<?>> futures = null;
+ if (numSplits > 1) {
+ futures = new ArrayList<>();
+ int numThreads = getConcurrencyLevel(numSplits) - 1;
+ cacheLoadTaskService = Executors.newFixedThreadPool(numThreads);
+ for (int i = 1; i < numSplits; i++) {
+ Future<?> future = cacheLoadTaskService.submit(cacheLoadTasks.get(i));
+ futures.add(future);
+ }
+ }
+ cacheLoadTasks.get(0).run();
Review Comment:
Why do we pick the first task to run synchronously instead of putting all tasks into the executor and wait for all futures to complete?
##########
flink-table/flink-table-runtime/src/main/java/org/apache/flink/table/runtime/keyselector/GenericRowDataKeySelector.java:
##########
@@ -0,0 +1,65 @@
+/*
+ * 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.flink.table.runtime.keyselector;
+
+import org.apache.flink.table.data.GenericRowData;
+import org.apache.flink.table.data.RowData;
+import org.apache.flink.table.runtime.generated.GeneratedProjection;
+import org.apache.flink.table.runtime.generated.Projection;
+import org.apache.flink.table.runtime.typeutils.InternalTypeInfo;
+import org.apache.flink.table.runtime.typeutils.RowDataSerializer;
+
+/** A KeySelector which will extract key from RowData. The key type is GenericRowData. */
+public class GenericRowDataKeySelector implements RowDataKeySelector {
+
+ private static final long serialVersionUID = 5375355285015381919L;
Review Comment:
`serialVersionUID` should start from 1
##########
flink-table/flink-table-runtime/src/main/java/org/apache/flink/table/runtime/functions/table/lookup/fullcache/LookupFullCache.java:
##########
@@ -0,0 +1,96 @@
+/*
+ * 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.flink.table.runtime.functions.table.lookup.fullcache;
+
+import org.apache.flink.configuration.Configuration;
+import org.apache.flink.metrics.groups.CacheMetricGroup;
+import org.apache.flink.table.connector.source.lookup.LookupOptions.LookupCacheType;
+import org.apache.flink.table.connector.source.lookup.cache.LookupCache;
+import org.apache.flink.table.connector.source.lookup.cache.trigger.CacheReloadTrigger;
+import org.apache.flink.table.data.RowData;
+import org.apache.flink.util.Preconditions;
+
+import java.util.Collection;
+import java.util.Collections;
+
+/** Internal implementation of {@link LookupCache} for {@link LookupCacheType#FULL}. */
+public class LookupFullCache implements LookupCache {
+ private static final long serialVersionUID = 1L;
+
+ private final CacheLoader cacheLoader;
+ private final CacheReloadTrigger reloadTrigger;
+
+ private volatile boolean metricsOpened;
+ private transient volatile ReloadTriggerContext reloadTriggerContext;
+ private transient volatile Throwable reloadFailCause;
+
+ public LookupFullCache(CacheLoader cacheLoader, CacheReloadTrigger reloadTrigger) {
+ this.cacheLoader = Preconditions.checkNotNull(cacheLoader);
+ this.reloadTrigger = Preconditions.checkNotNull(reloadTrigger);
+ }
+
+ @Override
+ public synchronized void open(CacheMetricGroup metricGroup) {
+ if (!metricsOpened) {
+ cacheLoader.open(metricGroup);
+ metricsOpened = true;
+ }
+ }
+
+ public synchronized void open(Configuration parameters) throws Exception {
+ if (reloadTriggerContext == null) {
+ cacheLoader.open(parameters);
+ reloadTriggerContext =
+ new ReloadTriggerContext(cacheLoader, th -> reloadFailCause = th);
Review Comment:
The exception consumer could be like storing the exception if `reloadFailCause` is null, otherwise `reloadFailCause.addSuppressed(th)`, to reveal all caught errors
##########
flink-table/flink-table-planner/src/main/java/org/apache/flink/table/planner/plan/utils/LookupJoinUtil.java:
##########
@@ -208,6 +217,21 @@ public static UserDefinedFunction getLookupFunction(
return new CachingAsyncLookupFunction(
partialCachingLookupProvider.getCache(),
partialCachingLookupProvider.createAsyncLookupFunction());
+ } else if (provider instanceof FullCachingLookupProvider) {
+ FullCachingLookupProvider fullCachingLookupProvider =
+ (FullCachingLookupProvider) provider;
+ RowType tableSourceRowType =
+ FlinkTypeFactory.toLogicalRowType(temporalTable.getRowType());
+ LookupFullCache fullCache =
+ classLoader == null
Review Comment:
Is there any specific reason for checking `classloader == null` ?
--
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: issues-unsubscribe@flink.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [flink] SmirAlex commented on a diff in pull request #20447: [FLINK-28532][table] Support full caching in lookup join runner using InputFormats as scan runtime provider
Posted by GitBox <gi...@apache.org>.
SmirAlex commented on code in PR #20447:
URL: https://github.com/apache/flink/pull/20447#discussion_r941496189
##########
flink-table/flink-table-runtime/src/main/java/org/apache/flink/table/runtime/functions/table/lookup/fullcache/inputformat/InputFormatCacheLoader.java:
##########
@@ -0,0 +1,159 @@
+/*
+ * 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.flink.table.runtime.functions.table.lookup.fullcache.inputformat;
+
+import org.apache.flink.api.common.io.InputFormat;
+import org.apache.flink.configuration.Configuration;
+import org.apache.flink.core.io.InputSplit;
+import org.apache.flink.table.data.RowData;
+import org.apache.flink.table.runtime.functions.table.lookup.fullcache.CacheLoader;
+import org.apache.flink.table.runtime.keyselector.GenericRowDataKeySelector;
+import org.apache.flink.table.runtime.typeutils.RowDataSerializer;
+import org.apache.flink.util.InstantiationUtil;
+import org.apache.flink.util.Preconditions;
+
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+
+import java.io.IOException;
+import java.util.ArrayList;
+import java.util.Arrays;
+import java.util.Collection;
+import java.util.List;
+import java.util.concurrent.ConcurrentHashMap;
+import java.util.concurrent.ExecutorService;
+import java.util.concurrent.Executors;
+import java.util.concurrent.Future;
+import java.util.stream.Collectors;
+
+/** {@link CacheLoader} that used {@link InputFormat} for loading data into the cache. */
+public class InputFormatCacheLoader extends CacheLoader {
+ private static final long serialVersionUID = 1L;
+ private static final Logger LOG = LoggerFactory.getLogger(InputFormatCacheLoader.class);
+
+ private final InputFormat<RowData, InputSplit> initialInputFormat;
+ private final GenericRowDataKeySelector keySelector;
+ private final RowDataSerializer cacheEntriesSerializer;
+
+ private transient volatile List<InputSplitCacheLoadTask> cacheLoadTasks;
+ private transient Configuration parameters;
+
+ private volatile boolean isStopped;
+
+ public InputFormatCacheLoader(
+ InputFormat<RowData, ?> initialInputFormat,
+ GenericRowDataKeySelector keySelector,
+ RowDataSerializer cacheEntriesSerializer) {
+ // noinspection unchecked
+ this.initialInputFormat = (InputFormat<RowData, InputSplit>) initialInputFormat;
+ this.keySelector = keySelector;
+ this.cacheEntriesSerializer = cacheEntriesSerializer;
+ }
+
+ @Override
+ public void open(Configuration parameters) throws Exception {
+ super.open(parameters);
+ this.parameters = parameters;
+ this.initialInputFormat.configure(parameters);
+ }
+
+ @Override
+ protected void reloadCache() throws Exception {
+ InputSplit[] inputSplits = createInputSplits();
+ int numSplits = inputSplits.length;
+ // load data into the another copy of cache
+ // notice: it requires twice more memory, but on the other hand we don't need any blocking
+ // cache has default initialCapacity and loadFactor, but overridden concurrencyLevel
+ ConcurrentHashMap<RowData, Collection<RowData>> newCache =
+ new ConcurrentHashMap<>(16, 0.75f, getConcurrencyLevel(numSplits));
+ this.cacheLoadTasks =
+ Arrays.stream(inputSplits)
+ .map(split -> createCacheLoadTask(split, newCache))
+ .collect(Collectors.toList());
+
+ // run first task and start numTasks - 1 threads to run remaining tasks
+ ExecutorService cacheLoadTaskService = null;
+ List<Future<?>> futures = null;
+ if (numSplits > 1) {
+ futures = new ArrayList<>();
+ int numThreads = getConcurrencyLevel(numSplits) - 1;
+ cacheLoadTaskService = Executors.newFixedThreadPool(numThreads);
+ for (int i = 1; i < numSplits; i++) {
+ Future<?> future = cacheLoadTaskService.submit(cacheLoadTasks.get(i));
+ futures.add(future);
+ }
+ }
+ cacheLoadTasks.get(0).run();
Review Comment:
By doing this we don't create one more thread. Parent thread any waits for all tasks to complete, so he can execute one of tasks to not idle.
--
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: issues-unsubscribe@flink.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [flink] PatrickRen commented on pull request #20447: [FLINK-28532][table] Support full caching in lookup join runner using InputFormats as scan runtime provider
Posted by GitBox <gi...@apache.org>.
PatrickRen commented on PR #20447:
URL: https://github.com/apache/flink/pull/20447#issuecomment-1209901001
An instance of green CI: https://dev.azure.com/renqs/Apache%20Flink/_build/results?buildId=389&view=results
--
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: issues-unsubscribe@flink.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [flink] flinkbot commented on pull request #20447: Flink 28532
Posted by GitBox <gi...@apache.org>.
flinkbot commented on PR #20447:
URL: https://github.com/apache/flink/pull/20447#issuecomment-1204770675
<!--
Meta data
{
"version" : 1,
"metaDataEntries" : [ {
"hash" : "76a2792d897bf3f6591ade52756bb8714259a963",
"status" : "UNKNOWN",
"url" : "TBD",
"triggerID" : "76a2792d897bf3f6591ade52756bb8714259a963",
"triggerType" : "PUSH"
} ]
}-->
## CI report:
* 76a2792d897bf3f6591ade52756bb8714259a963 UNKNOWN
<details>
<summary>Bot commands</summary>
The @flinkbot bot supports the following commands:
- `@flinkbot run azure` re-run the last Azure build
</details>
--
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: issues-unsubscribe@flink.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [flink] SmirAlex commented on a diff in pull request #20447: [FLINK-28532][table] Support full caching in lookup join runner using InputFormats as scan runtime provider
Posted by GitBox <gi...@apache.org>.
SmirAlex commented on code in PR #20447:
URL: https://github.com/apache/flink/pull/20447#discussion_r941496189
##########
flink-table/flink-table-runtime/src/main/java/org/apache/flink/table/runtime/functions/table/lookup/fullcache/inputformat/InputFormatCacheLoader.java:
##########
@@ -0,0 +1,159 @@
+/*
+ * 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.flink.table.runtime.functions.table.lookup.fullcache.inputformat;
+
+import org.apache.flink.api.common.io.InputFormat;
+import org.apache.flink.configuration.Configuration;
+import org.apache.flink.core.io.InputSplit;
+import org.apache.flink.table.data.RowData;
+import org.apache.flink.table.runtime.functions.table.lookup.fullcache.CacheLoader;
+import org.apache.flink.table.runtime.keyselector.GenericRowDataKeySelector;
+import org.apache.flink.table.runtime.typeutils.RowDataSerializer;
+import org.apache.flink.util.InstantiationUtil;
+import org.apache.flink.util.Preconditions;
+
+import org.slf4j.Logger;
+import org.slf4j.LoggerFactory;
+
+import java.io.IOException;
+import java.util.ArrayList;
+import java.util.Arrays;
+import java.util.Collection;
+import java.util.List;
+import java.util.concurrent.ConcurrentHashMap;
+import java.util.concurrent.ExecutorService;
+import java.util.concurrent.Executors;
+import java.util.concurrent.Future;
+import java.util.stream.Collectors;
+
+/** {@link CacheLoader} that used {@link InputFormat} for loading data into the cache. */
+public class InputFormatCacheLoader extends CacheLoader {
+ private static final long serialVersionUID = 1L;
+ private static final Logger LOG = LoggerFactory.getLogger(InputFormatCacheLoader.class);
+
+ private final InputFormat<RowData, InputSplit> initialInputFormat;
+ private final GenericRowDataKeySelector keySelector;
+ private final RowDataSerializer cacheEntriesSerializer;
+
+ private transient volatile List<InputSplitCacheLoadTask> cacheLoadTasks;
+ private transient Configuration parameters;
+
+ private volatile boolean isStopped;
+
+ public InputFormatCacheLoader(
+ InputFormat<RowData, ?> initialInputFormat,
+ GenericRowDataKeySelector keySelector,
+ RowDataSerializer cacheEntriesSerializer) {
+ // noinspection unchecked
+ this.initialInputFormat = (InputFormat<RowData, InputSplit>) initialInputFormat;
+ this.keySelector = keySelector;
+ this.cacheEntriesSerializer = cacheEntriesSerializer;
+ }
+
+ @Override
+ public void open(Configuration parameters) throws Exception {
+ super.open(parameters);
+ this.parameters = parameters;
+ this.initialInputFormat.configure(parameters);
+ }
+
+ @Override
+ protected void reloadCache() throws Exception {
+ InputSplit[] inputSplits = createInputSplits();
+ int numSplits = inputSplits.length;
+ // load data into the another copy of cache
+ // notice: it requires twice more memory, but on the other hand we don't need any blocking
+ // cache has default initialCapacity and loadFactor, but overridden concurrencyLevel
+ ConcurrentHashMap<RowData, Collection<RowData>> newCache =
+ new ConcurrentHashMap<>(16, 0.75f, getConcurrencyLevel(numSplits));
+ this.cacheLoadTasks =
+ Arrays.stream(inputSplits)
+ .map(split -> createCacheLoadTask(split, newCache))
+ .collect(Collectors.toList());
+
+ // run first task and start numTasks - 1 threads to run remaining tasks
+ ExecutorService cacheLoadTaskService = null;
+ List<Future<?>> futures = null;
+ if (numSplits > 1) {
+ futures = new ArrayList<>();
+ int numThreads = getConcurrencyLevel(numSplits) - 1;
+ cacheLoadTaskService = Executors.newFixedThreadPool(numThreads);
+ for (int i = 1; i < numSplits; i++) {
+ Future<?> future = cacheLoadTaskService.submit(cacheLoadTasks.get(i));
+ futures.add(future);
+ }
+ }
+ cacheLoadTasks.get(0).run();
Review Comment:
By doing this we don't create one more thread. Parent thread anyway waits for all tasks to complete, so it can execute one of tasks to not idle.
--
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: issues-unsubscribe@flink.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
[GitHub] [flink] SmirAlex commented on pull request #20447: [FLINK-28532][table] Support full caching in lookup join runner using InputFormats as scan runtime provider
Posted by GitBox <gi...@apache.org>.
SmirAlex commented on PR #20447:
URL: https://github.com/apache/flink/pull/20447#issuecomment-1209576777
@PatrickRen Thanks for the review! Please look at latest commit where I addressed your comments
--
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: issues-unsubscribe@flink.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org