[
https://issues.apache.org/jira/browse/HDDS-2203?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16943445#comment-16943445
]
Marton Elek commented on HDDS-2203:
-----------------------------------
I would be happy to commit this patch as is. I think it's good to have as we
have this problem in the code. (Don't know what is the timeline for the
refactor, but for me it seems to be reasonable to commit this until the
refactor provides a more permanent solution).
> Race condition in ByteStringHelper.init()
> -----------------------------------------
>
> Key: HDDS-2203
> URL: https://issues.apache.org/jira/browse/HDDS-2203
> Project: Hadoop Distributed Data Store
> Issue Type: Bug
> Components: Ozone Client, SCM
> Reporter: Istvan Fajth
> Assignee: Istvan Fajth
> Priority: Critical
> Time Spent: 1h 40m
> Remaining Estimate: 0h
>
> The current init method:
> {code}
> public static void init(boolean isUnsafeByteOperation) {
> final boolean set = INITIALIZED.compareAndSet(false, true);
> if (set) {
> ByteStringHelper.isUnsafeByteOperationsEnabled =
> isUnsafeByteOperation;
> } else {
> // already initialized, check values
> Preconditions.checkState(isUnsafeByteOperationsEnabled
> == isUnsafeByteOperation);
> }
> }
> {code}
> In a scenario when two thread accesses this method, and the execution order
> is the following, then the second thread runs into an exception from
> PreCondition.checkState() in the else branch.
> In an unitialized state:
> - T1 thread arrives to the method with true as the parameter, the class
> initialises the isUnsafeByteOperationsEnabled to false
> - T1 sets INITIALIZED true
> - T2 arrives to the method with true as the parameter
> - T2 reads the INITALIZED value and as it is not false goes to else branch
> - T2 tries to check if the internal boolean property is the same true as it
> wanted to set, and as T1 still to set the value, the checkState throws an
> IllegalArgumentException.
> This happens in certain Hive query cases, as it came from that testing, the
> exception we see there is the following:
> {code}
> Error: Error while processing statement: FAILED: Execution Error, return code
> 2 from org.apache.hadoop.hive.ql.exec.tez.TezTask. Vertex failed,
> vertexName=Map 2, vertexId=vertex_1569486223160_0334_1_02,
> diagnostics=[Vertex vertex_1569486223160_0334_1_02 [Map 2] killed/failed
> due to:ROOT_INPUT_INIT_FAILURE, Vertex Input: item initializer failed,
> vertex=vertex_1569486223160_0334_1_02 [Map 2], java.io.IOException: Couldn't
> create RpcClient protocol
> at
> org.apache.hadoop.ozone.client.OzoneClientFactory.getClientProtocol(OzoneClientFactory.java:263)
> at
> org.apache.hadoop.ozone.client.OzoneClientFactory.getClientProtocol(OzoneClientFactory.java:239)
> at
> org.apache.hadoop.ozone.client.OzoneClientFactory.getRpcClient(OzoneClientFactory.java:203)
> at
> org.apache.hadoop.ozone.client.OzoneClientFactory.getRpcClient(OzoneClientFactory.java:165)
> at
> org.apache.hadoop.fs.ozone.BasicOzoneClientAdapterImpl.<init>(BasicOzoneClientAdapterImpl.java:158)
> at
> org.apache.hadoop.fs.ozone.OzoneClientAdapterImpl.<init>(OzoneClientAdapterImpl.java:50)
> at
> org.apache.hadoop.fs.ozone.OzoneFileSystem.createAdapter(OzoneFileSystem.java:102)
> at
> org.apache.hadoop.fs.ozone.BasicOzoneFileSystem.initialize(BasicOzoneFileSystem.java:155)
> at org.apache.hadoop.fs.FileSystem.createFileSystem(FileSystem.java:3315)
> at org.apache.hadoop.fs.FileSystem.access$200(FileSystem.java:136)
> at org.apache.hadoop.fs.FileSystem$Cache.getInternal(FileSystem.java:3364)
> at org.apache.hadoop.fs.FileSystem$Cache.get(FileSystem.java:3332)
> at org.apache.hadoop.fs.FileSystem.get(FileSystem.java:491)
> at org.apache.hadoop.fs.Path.getFileSystem(Path.java:361)
> at
> org.apache.hadoop.hive.ql.io.orc.OrcInputFormat.generateSplitsInfo(OrcInputFormat.java:1821)
> at
> org.apache.hadoop.hive.ql.io.orc.OrcInputFormat.getSplits(OrcInputFormat.java:2002)
> at
> org.apache.hadoop.hive.ql.io.HiveInputFormat.addSplitsForGroup(HiveInputFormat.java:524)
> at
> org.apache.hadoop.hive.ql.io.HiveInputFormat.getSplits(HiveInputFormat.java:781)
> at
> org.apache.hadoop.hive.ql.exec.tez.HiveSplitGenerator.initialize(HiveSplitGenerator.java:243)
> at
> org.apache.tez.dag.app.dag.RootInputInitializerManager$InputInitializerCallable$1.run(RootInputInitializerManager.java:278)
> at
> org.apache.tez.dag.app.dag.RootInputInitializerManager$InputInitializerCallable$1.run(RootInputInitializerManager.java:269)
> at java.security.AccessController.doPrivileged(Native Method)
> at javax.security.auth.Subject.doAs(Subject.java:422)
> at
> org.apache.hadoop.security.UserGroupInformation.doAs(UserGroupInformation.java:1876)
> at
> org.apache.tez.dag.app.dag.RootInputInitializerManager$InputInitializerCallable.call(RootInputInitializerManager.java:269)
> at
> org.apache.tez.dag.app.dag.RootInputInitializerManager$InputInitializerCallable.call(RootInputInitializerManager.java:253)
> at
> com.google.common.util.concurrent.TrustedListenableFutureTask$TrustedFutureInterruptibleTask.runInterruptibly(TrustedListenableFutureTask.java:108)
> at
> com.google.common.util.concurrent.InterruptibleTask.run(InterruptibleTask.java:41)
> at
> com.google.common.util.concurrent.TrustedListenableFutureTask.run(TrustedListenableFutureTask.java:77)
> at
> java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
> at
> java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
> at java.lang.Thread.run(Thread.java:748)
> Caused by: java.lang.IllegalStateException
> at
> org.apache.hadoop.ozone.shaded.com.google.common.base.Preconditions.checkState(Preconditions.java:129)
> at
> org.apache.hadoop.hdds.scm.ByteStringHelper.init(ByteStringHelper.java:47)
> at org.apache.hadoop.ozone.client.rpc.RpcClient.<init>(RpcClient.java:241)
> at
> org.apache.hadoop.ozone.client.OzoneClientFactory.getClientProtocol(OzoneClientFactory.java:256)
> ... 31 more
> {code}
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: hdfs-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: hdfs-issues-h...@hadoop.apache.org
