HBASE-24896 'Stuck' in static initialization creating RegionInfo instance

Closes #2423

Untangle RegionInfo, RegionInfoBuilder, and MutableRegionInfo static
initializations some. Move MutableRegionInfo from inner-class of
RegionInfoBuilder to be (package private) standalone. Undo static
initializing references from RI to RIB.

Co-authored-by: Nick Dimiduk <ndimiduk@apache.org>
Signed-off-by: Bharath Vissapragada <bharathv@apache.org>
Signed-off-by: Duo Zhang <zhangduo@apache.org>
Signed-off-by: Viraj Jasani <vjasani@apache.org>
This commit is contained in:
stack 2020-09-20 14:08:45 +05:30 committed by Viraj Jasani
parent fe03a35ea2
commit cef1220590
No known key found for this signature in database
GPG Key ID: B3D6C0B41C8ADFD5
2 changed files with 77 additions and 3 deletions

View File

@ -71,7 +71,9 @@ public interface RegionInfo extends Comparable<RegionInfo> {
*/
@Deprecated
@InterfaceAudience.Private
RegionInfo UNDEFINED = RegionInfoBuilder.newBuilder(TableName.valueOf("__UNDEFINED__")).build();
// Not using RegionInfoBuilder intentionally to avoid a static loading deadlock: HBASE-24627
RegionInfo UNDEFINED = new MutableRegionInfo(0, TableName.valueOf("__UNDEFINED__"),
RegionInfo.DEFAULT_REPLICA_ID);
/**
* Separator used to demarcate the encodedName in a region name
@ -588,8 +590,9 @@ public interface RegionInfo extends Comparable<RegionInfo> {
* @return the MOB {@link RegionInfo}.
*/
static RegionInfo createMobRegionInfo(TableName tableName) {
return RegionInfoBuilder.newBuilder(tableName).setStartKey(Bytes.toBytes(".mob")).
setRegionId(0).build();
// Skipping reference to RegionInfoBuilder in this class.
return new MutableRegionInfo(tableName, Bytes.toBytes(".mob"),
HConstants.EMPTY_END_ROW, false, 0, DEFAULT_REPLICA_ID, false);
}
/**

View File

@ -0,0 +1,71 @@
/*
* 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.hadoop.hbase.regionserver;
import java.util.concurrent.CompletableFuture;
import java.util.concurrent.TimeUnit;
import java.util.function.Supplier;
import java.util.stream.Stream;
import org.apache.hadoop.hbase.HBaseClassTestRule;
import org.apache.hadoop.hbase.HRegionInfo;
import org.apache.hadoop.hbase.client.RegionInfo;
import org.apache.hadoop.hbase.client.RegionInfoBuilder;
import org.apache.hadoop.hbase.testclassification.LargeTests;
import org.apache.hadoop.hbase.testclassification.RegionServerTests;
import org.junit.ClassRule;
import org.junit.Test;
import org.junit.experimental.categories.Category;
/**
* Test for the tangled mess that is static initialization of our our {@link HRegionInfo} and
* {@link RegionInfoBuilder}, as reported on HBASE-24896. The condition being tested can only be
* reproduced the first time a JVM loads the classes under test. Thus, this test is marked as a
* {@link LargeTests} because, under their current configuration, tests in that category are run
* in their own JVM instances.
*/
@SuppressWarnings("deprecation")
@Category({ RegionServerTests.class, LargeTests.class})
public class TestRegionInfoStaticInitialization {
@ClassRule
public static final HBaseClassTestRule CLASS_RULE =
HBaseClassTestRule.forClass(TestRegionInfoStaticInitialization.class);
@Test
public void testParallelStaticInitialization() throws Exception {
// The JVM loads symbols lazily. These suppliers reference two symbols that, before this patch,
// are mutually dependent and expose a deadlock in the loading of symbols from RegionInfo and
// RegionInfoBuilder.
final Supplier<RegionInfo> retrieveUNDEFINED = () -> HRegionInfo.UNDEFINED;
final Supplier<RegionInfo> retrieveMetaRegionInfo =
() -> RegionInfoBuilder.FIRST_META_REGIONINFO;
// The test runs multiple threads that reference these mutually dependent symbols. In order to
// express this bug, these threads need to access these symbols at roughly the same time, so
// that the classloader is asked to materialize these symbols concurrently. These Suppliers are
// run on threads that have already been allocated, managed by the system's ForkJoin pool.
final CompletableFuture<?>[] futures = Stream.of(
retrieveUNDEFINED, retrieveMetaRegionInfo, retrieveUNDEFINED, retrieveMetaRegionInfo)
.map(CompletableFuture::supplyAsync)
.toArray(CompletableFuture<?>[]::new);
// Loading classes should be relatively fast. 5 seconds is an arbitrary choice of timeout. It
// was chosen under the assumption that loading these symbols should complete much faster than
// this window.
CompletableFuture.allOf(futures).get(5, TimeUnit.SECONDS);
}
}