[
https://issues.apache.org/jira/browse/FLINK-30075?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17635770#comment-17635770
]
zhanglu153 edited comment on FLINK-30075 at 11/19/22 6:05 AM:
--------------------------------------------------------------
This error is related to the hadoop version and the operation of setting
ChildFirstClassLoader to null. The main reason is that the static code block in
class org.apache.hadoop.yarn.conf.YarnConfiguration sets the properties object
in the Configuration object to null.
The solutions considered are as follows:
# Set classloader.check-leaked-classloader option is false, but it will cause
leakage of the Flink's ChildFirstClassLoader.
# Set the fs.hdfs.impl.disable.cache parameter to true in the default
configuration file of hadoop. It is equivalent to creating a new filesystem
file handle every time. For tasks, a new file handle will be created when
creating new splits. It needs to be closed in time, otherwise it will lead to
the task oom or other situations.
# Modify method
org.apache.flink.connectors.hive.read.HiveTableInputFormat#createInputSplits(int,
java.util.List<org.apache.flink.connectors.hive.HiveTablePartition>,
org.apache.hadoop.mapred.JobConf) and create a YarnConfiguration object before
calling org.apache.hadoop.fs.FileSystem#exists method, as shown below:
{code:java}
public static HiveTableInputSplit[] createInputSplits(
int minNumSplits, List<HiveTablePartition> partitions, JobConf jobConf)
throws IOException {
YarnConfiguration yarnConfiguration = new YarnConfiguration();
List<HiveTableInputSplit> hiveSplits = new ArrayList<>();
int splitNum = 0;
FileSystem fs = null;
for (HiveTablePartition partition : partitions) {
StorageDescriptor sd = partition.getStorageDescriptor();
Path inputPath = new Path(sd.getLocation());
if (fs == null) {
fs = inputPath.getFileSystem(jobConf);
}
// it's possible a partition exists in metastore but the data has been
removed
if (!fs.exists(inputPath)) {
continue;
}
InputFormat format;
try {
format =
(InputFormat)
Class.forName(
sd.getInputFormat(),
true,
Thread.currentThread().getContextClassLoader())
.newInstance();
} catch (Exception e) {
throw new FlinkHiveException("Unable to instantiate the hadoop
input format", e);
}
ReflectionUtils.setConf(format, jobConf);
jobConf.set(INPUT_DIR, sd.getLocation());
// TODO: we should consider how to calculate the splits according to
minNumSplits in the
// future.
org.apache.hadoop.mapred.InputSplit[] splitArray =
format.getSplits(jobConf, minNumSplits);
for (org.apache.hadoop.mapred.InputSplit inputSplit : splitArray) {
hiveSplits.add(new HiveTableInputSplit(splitNum++, inputSplit,
jobConf, partition));
}
} {code}
In this way, the static code block in YarnConfiguration can be executed in
advance when the YarnConfiguration class is loaded for the first time. This can
avoid setting the properties object in the Configuration object to null when
calling the org.apache.hadoop.mapred.FileInputFormat#getSplits method. This
method seems to actually solve this problem, but it seems unfriendly, because
the YarnConfiguration object is not needed to use in the actual
org.apache.flink.connectors.hive.read.HiveTableInputFormat class.
was (Author: JIRAUSER298392):
This error is related to the hadoop version and the operation of setting
ChildFirstClassLoader to null. The main reason is that the static code block in
class org.apache.hadoop.yarn.conf.YarnConfiguration sets the properties object
in the Configuration object to null.
The solutions considered are as follows:
# Set classloader.check-leaked-classloader option is false, but it will cause
leakage of the Flink's ChildFirstClassLoader.
# Set the fs.hdfs.impl.disable.cache parameter to true in the default
configuration file of hadoop. It is equivalent to creating a new filesystem
file handle every time. For tasks, a new file handle will be created when
creating new splits. It needs to be closed in time, otherwise it will lead to
the task oom or other situations.
# Modify method
org.apache.flink.connectors.hive.read.HiveTableInputFormat#createInputSplits(int,
java.util.List<org.apache.flink.connectors.hive.HiveTablePartition>,
org.apache.hadoop.mapred.JobConf) and create a YarnConfiguration object before
calling org.apache.hadoop.fs.FileSystem#exists method, as shown below:
{code:java}
public static HiveTableInputSplit[] createInputSplits(
int minNumSplits, List<HiveTablePartition> partitions, JobConf jobConf)
throws IOException {
YarnConfiguration yarnConfiguration = new YarnConfiguration();
List<HiveTableInputSplit> hiveSplits = new ArrayList<>();
int splitNum = 0;
FileSystem fs = null;
for (HiveTablePartition partition : partitions) {
StorageDescriptor sd = partition.getStorageDescriptor();
Path inputPath = new Path(sd.getLocation());
if (fs == null) {
fs = inputPath.getFileSystem(jobConf);
}
// it's possible a partition exists in metastore but the data has been
removed
if (!fs.exists(inputPath)) {
continue;
}
InputFormat format;
try {
format =
(InputFormat)
Class.forName(
sd.getInputFormat(),
true,
Thread.currentThread().getContextClassLoader())
.newInstance();
} catch (Exception e) {
throw new FlinkHiveException("Unable to instantiate the hadoop
input format", e);
}
ReflectionUtils.setConf(format, jobConf);
jobConf.set(INPUT_DIR, sd.getLocation());
// TODO: we should consider how to calculate the splits according to
minNumSplits in the
// future.
org.apache.hadoop.mapred.InputSplit[] splitArray =
format.getSplits(jobConf, minNumSplits);
for (org.apache.hadoop.mapred.InputSplit inputSplit : splitArray) {
hiveSplits.add(new HiveTableInputSplit(splitNum++, inputSplit,
jobConf, partition));
}
} {code}
In this way, the static code block in YarnConfiguration can be executed in
advance when the YarnConfiguration class is loaded for the first time. This can
avoid setting the properties object in the Configuration object to null when
calling the org.apache.hadoop.mapred.FileInputFormat#getSplits method. This
method seems to actually solve this problem, but it seems unfriendly, because
the YarnConfiguration object is not needed in the actual
org.apache.flink.connectors.hive.read.HiveTableInputFormat class.
> Failed to load data to the cache after the hive lookup join task is restarted
> -----------------------------------------------------------------------------
>
> Key: FLINK-30075
> URL: https://issues.apache.org/jira/browse/FLINK-30075
> Project: Flink
> Issue Type: Bug
> Components: Connectors / Hive
> Affects Versions: 1.12.2
> Reporter: zhanglu153
> Priority: Major
> Attachments: Set the properties object in the Configuration object of
> DFSClient#namenode proxy in the DistributedFileSystem object to null.png, The
> error thrown when calling the getProps method after properties is set to
> null.png, call the org.apache.hadoop.conf.Configuration#getProps method to
> create the Properties object.png, hadoop2.10.1
> org.apache.hadoop.conf.Configuration#getProps.png, hadoop2.10.1
> org.apache.hadoop.mapred.Master#getMasterPrincipal.png, hadoop2.10.1
> org.apache.hadoop.security.SaslRpcClient#conf.png, hadoop2.10.1
> org.apache.hadoop.yarn.client.util.YarnClientUtils#getYarnConfWithRmHaId.png,
> hadoop2.10.1 org.apache.hadoop.yarn.conf.YarnConfiguration static code
> block.png, hadoop2.8.5
> org.apache.hadoop.mapred.Master#getMasterPrincipal.png, hadoop2.8.5
> org.apache.hadoop.security.SaslRpcClient#conf.png,
> org.apache.flink.connectors.hive.read.HiveTableInputFormat#createInputSplits.png,
> org.apache.hadoop.conf.Configuration#addDefaultResource.png,
> org.apache.hadoop.security.SaslRpcClient.png
>
>
> When I test kafka and hive lookup join, only the connection between tm and
> zookeeper is disconnected after the data is successfully loaded into the
> cache. In this case, the task will be restarted because the flink task
> restart policy is configured (note that the task process does not change at
> this time, and the task restarts in the same container).
> The task sql is as follows:
> {code:java}
> StreamExecutionEnvironment env =
> StreamExecutionEnvironment.getExecutionEnvironment();
> env.setParallelism(1);
> EnvironmentSettings settings =
> EnvironmentSettings.newInstance().useBlinkPlanner().inStreamingMode().build();
> StreamTableEnvironment tableEnv = StreamTableEnvironment.create(env,
> settings);
> tableEnv.executeSql("CREATE TABLE kafka_table_1 (\n" +
> " a string,\n" +
> " b int,\n" +
> " c string,\n" +
> " log_ts TIMESTAMP(0),\n" +
> ") WITH (\n" +
> " 'connector' = 'kafka',\n" +
> " 'topic' = 'test',\n" +
> " 'scan.startup.mode' = 'latest-offset',\n" +
> " 'properties.bootstrap.servers' = '172.16.144.208:9092',\n" +
> " 'format' = 'csv'\n" +
> ")");
> String name = "hive_catalog";
> String hiveConfDir = "/cloud/service/flink/conf";
> HiveCatalog hive = new HiveCatalog(name, null, hiveConfDir);
> hive.open();
> tableEnv.registerCatalog(name, hive);
> tableEnv.useCatalog(name);
> tableEnv.getConfig().setSqlDialect(SqlDialect.HIVE);
> tableEnv.executeSql("create table if not exists ttt(a string,b int,c
> string)");
> tableEnv.getConfig().setSqlDialect(SqlDialect.DEFAULT);
> tableEnv.executeSql("select * from
> default_catalog.default_database.kafka_table_1 as u join ttt for system_time
> as of u.proctime as o on u.a = o.a").print();{code}
> After the task is successfully restarted, kafka produces data. When some data
> comes in, an error message is displayed indicating that loading data into the
> cache fails. The error message is as follows:
> {code:java}
> 2022-11-17 14:09:16,041 INFO
> org.apache.flink.table.filesystem.FileSystemLookupFunction [] - Populating
> lookup join cache
> 2022-11-17 14:09:16,051 INFO org.apache.hadoop.conf.Configuration
> [] - getProps 1230612033
> 2022-11-17 14:09:16,053 INFO
> org.apache.hadoop.io.retry.RetryInvocationHandler [] -
> java.io.IOException:
> org.apache.flink.shaded.hadoop2.com.google.protobuf.ServiceException:
> java.lang.IllegalStateException: Trying to access closed classloader. Please
> check if you store classloaders directly or indirectly in static fields. If
> the stacktrace suggests that the leak occurs in a third party library and
> cannot be fixed immediately, you can disable this check with the
> configuration 'classloader.check-leaked-classloader'., while invoking
> ClientNamenodeProtocolTranslatorPB.getFileInfo over
> hdp-hadoop-hdp-namenode-1.hdp-hadoop-hdp-namenode.hdp-dev-env-3.svc.cluster.local/192.168.6.85:9000.
> Trying to failover immediately.
> 2022-11-17 14:09:16,053 DEBUG org.apache.hadoop.ipc.Client
> [] - The ping interval is 60000 ms.
> 2022-11-17 14:09:16,053 DEBUG org.apache.hadoop.ipc.Client
> [] - Connecting to
> hdp-hadoop-hdp-namenode-0.hdp-hadoop-hdp-namenode.hdp-dev-env-3.svc.cluster.local/192.168.0.8:9000
> 2022-11-17 14:09:16,055 DEBUG org.apache.hadoop.security.UserGroupInformation
> [] - PrivilegedAction
> as:hadoop/hdp-flinkhistory-hdp-flink-history-6754db4f9c-8b8nd.hdp-flinkhistory-hdp-flink-history.hdp-dev-env-3.svc.cluster.lo...@dahua.com
> (auth:KERBEROS)
> from:org.apache.hadoop.ipc.Client$Connection.setupIOstreams(Client.java:825)
> 2022-11-17 14:09:16,056 DEBUG org.apache.hadoop.security.SaslRpcClient
> [] - Sending sasl message state: NEGOTIATE
> 2022-11-17 14:09:16,057 DEBUG org.apache.hadoop.security.SaslRpcClient
> [] - Get token info proto:interface
> org.apache.hadoop.hdfs.protocolPB.ClientNamenodeProtocolPB
> info:@org.apache.hadoop.security.token.TokenInfo(value=class
> org.apache.hadoop.hdfs.security.token.delegation.DelegationTokenSelector)
> 2022-11-17 14:09:16,057 INFO org.apache.hadoop.security.SaslRpcClient
> [] - getServerPrincipal 1230612033
> 2022-11-17 14:09:16,057 DEBUG org.apache.hadoop.security.SaslRpcClient
> [] - Get kerberos info proto:interface
> org.apache.hadoop.hdfs.protocolPB.ClientNamenodeProtocolPB
> info:@org.apache.hadoop.security.KerberosInfo(clientPrincipal=,
> serverPrincipal=dfs.namenode.kerberos.principal)
> 2022-11-17 14:09:16,059 DEBUG org.apache.hadoop.security.SaslRpcClient
> [] - getting serverKey: dfs.namenode.kerberos.principal conf
> value: null principal: null
> 2022-11-17 14:09:16,059 DEBUG org.apache.hadoop.ipc.Client
> [] - closing ipc connection to
> hdp-hadoop-hdp-namenode-0.hdp-hadoop-hdp-namenode.hdp-dev-env-3.svc.cluster.local/192.168.0.8:9000:
> Couldn't set up IO streams: java.lang.IllegalArgumentException: Failed to
> specify server's Kerberos principal name
> java.io.IOException: Couldn't set up IO streams:
> java.lang.IllegalArgumentException: Failed to specify server's Kerberos
> principal name
> at
> org.apache.hadoop.ipc.Client$Connection.setupIOstreams(Client.java:891)
> ~[flink-shaded-hadoop-2-uber-2.10.1-HDP-22.09.1-release-10.0.jar:2.10.1-HDP-22.09.1-release-10.0]
> at org.apache.hadoop.ipc.Client$Connection.access$3700(Client.java:423)
> ~[flink-shaded-hadoop-2-uber-2.10.1-HDP-22.09.1-release-10.0.jar:2.10.1-HDP-22.09.1-release-10.0]
> at org.apache.hadoop.ipc.Client.getConnection(Client.java:1615)
> ~[flink-shaded-hadoop-2-uber-2.10.1-HDP-22.09.1-release-10.0.jar:2.10.1-HDP-22.09.1-release-10.0]
> at org.apache.hadoop.ipc.Client.call(Client.java:1446)
> ~[flink-shaded-hadoop-2-uber-2.10.1-HDP-22.09.1-release-10.0.jar:2.10.1-HDP-22.09.1-release-10.0]
> at org.apache.hadoop.ipc.Client.call(Client.java:1399)
> ~[flink-shaded-hadoop-2-uber-2.10.1-HDP-22.09.1-release-10.0.jar:2.10.1-HDP-22.09.1-release-10.0]
> at
> org.apache.hadoop.ipc.ProtobufRpcEngine$Invoker.invoke(ProtobufRpcEngine.java:232)
>
> ~[flink-shaded-hadoop-2-uber-2.10.1-HDP-22.09.1-release-10.0.jar:2.10.1-HDP-22.09.1-release-10.0]
> at
> org.apache.hadoop.ipc.ProtobufRpcEngine$Invoker.invoke(ProtobufRpcEngine.java:118)
>
> ~[flink-shaded-hadoop-2-uber-2.10.1-HDP-22.09.1-release-10.0.jar:2.10.1-HDP-22.09.1-release-10.0]
> at com.sun.proxy.$Proxy27.getFileInfo(Unknown Source) ~[?:?]
> at
> org.apache.hadoop.hdfs.protocolPB.ClientNamenodeProtocolTranslatorPB.getFileInfo(ClientNamenodeProtocolTranslatorPB.java:800)
>
> ~[flink-shaded-hadoop-2-uber-2.10.1-HDP-22.09.1-release-10.0.jar:2.10.1-HDP-22.09.1-release-10.0]
> at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
> ~[?:1.8.0_342]
> at
> sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:62)
> ~[?:1.8.0_342]
> at
> sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
> ~[?:1.8.0_342]
> at java.lang.reflect.Method.invoke(Method.java:498) ~[?:1.8.0_342]
> at
> org.apache.hadoop.io.retry.RetryInvocationHandler.invokeMethod(RetryInvocationHandler.java:422)
>
> ~[flink-shaded-hadoop-2-uber-2.10.1-HDP-22.09.1-release-10.0.jar:2.10.1-HDP-22.09.1-release-10.0]
> at
> org.apache.hadoop.io.retry.RetryInvocationHandler$Call.invokeMethod(RetryInvocationHandler.java:165)
>
> ~[flink-shaded-hadoop-2-uber-2.10.1-HDP-22.09.1-release-10.0.jar:2.10.1-HDP-22.09.1-release-10.0]
> at
> org.apache.hadoop.io.retry.RetryInvocationHandler$Call.invoke(RetryInvocationHandler.java:157)
>
> ~[flink-shaded-hadoop-2-uber-2.10.1-HDP-22.09.1-release-10.0.jar:2.10.1-HDP-22.09.1-release-10.0]
> at
> org.apache.hadoop.io.retry.RetryInvocationHandler$Call.invokeOnce(RetryInvocationHandler.java:95)
>
> ~[flink-shaded-hadoop-2-uber-2.10.1-HDP-22.09.1-release-10.0.jar:2.10.1-HDP-22.09.1-release-10.0]
> at
> org.apache.hadoop.io.retry.RetryInvocationHandler.invoke(RetryInvocationHandler.java:359)
>
> ~[flink-shaded-hadoop-2-uber-2.10.1-HDP-22.09.1-release-10.0.jar:2.10.1-HDP-22.09.1-release-10.0]
> at com.sun.proxy.$Proxy28.getFileInfo(Unknown Source) ~[?:?]
> at org.apache.hadoop.hdfs.DFSClient.getFileInfo(DFSClient.java:1673)
> ~[flink-shaded-hadoop-2-uber-2.10.1-HDP-22.09.1-release-10.0.jar:2.10.1-HDP-22.09.1-release-10.0]
> at
> org.apache.hadoop.hdfs.DistributedFileSystem$29.doCall(DistributedFileSystem.java:1524)
>
> ~[flink-shaded-hadoop-2-uber-2.10.1-HDP-22.09.1-release-10.0.jar:2.10.1-HDP-22.09.1-release-10.0]
> at
> org.apache.hadoop.hdfs.DistributedFileSystem$29.doCall(DistributedFileSystem.java:1521)
>
> ~[flink-shaded-hadoop-2-uber-2.10.1-HDP-22.09.1-release-10.0.jar:2.10.1-HDP-22.09.1-release-10.0]
> at
> org.apache.hadoop.fs.FileSystemLinkResolver.resolve(FileSystemLinkResolver.java:81)
>
> ~[flink-shaded-hadoop-2-uber-2.10.1-HDP-22.09.1-release-10.0.jar:2.10.1-HDP-22.09.1-release-10.0]
> at
> org.apache.hadoop.hdfs.DistributedFileSystem.getFileStatus(DistributedFileSystem.java:1536)
>
> ~[flink-shaded-hadoop-2-uber-2.10.1-HDP-22.09.1-release-10.0.jar:2.10.1-HDP-22.09.1-release-10.0]
> at org.apache.hadoop.fs.FileSystem.exists(FileSystem.java:1632)
> ~[flink-shaded-hadoop-2-uber-2.10.1-HDP-22.09.1-release-10.0.jar:2.10.1-HDP-22.09.1-release-10.0]
> at
> org.apache.flink.connectors.hive.read.HiveTableInputFormat.createInputSplits(HiveTableInputFormat.java:334)
> ~[flink-connector-hive_2.11-1.12.2-HDP-22.10.28.jar:1.12.2-HDP-22.10.28]
> at
> org.apache.flink.connectors.hive.read.HiveTableInputFormat.createInputSplits(HiveTableInputFormat.java:318)
> ~[flink-connector-hive_2.11-1.12.2-HDP-22.10.28.jar:1.12.2-HDP-22.10.28]
> at
> org.apache.flink.connectors.hive.read.HiveInputFormatPartitionReader.open(HiveInputFormatPartitionReader.java:83)
> ~[flink-connector-hive_2.11-1.12.2-HDP-22.10.28.jar:1.12.2-HDP-22.10.28]
> at
> org.apache.flink.table.filesystem.FileSystemLookupFunction.checkCacheReload(FileSystemLookupFunction.java:132)
> ~[flink-table-blink_2.11-1.12.2-HDP-22.10.09.jar:1.12.2-HDP-22.10.09]
> at
> org.apache.flink.table.filesystem.FileSystemLookupFunction.eval(FileSystemLookupFunction.java:105)
> ~[flink-table-blink_2.11-1.12.2-HDP-22.10.09.jar:1.12.2-HDP-22.10.09]
> at
> org.apache.flink.table.filesystem.security.HdpFileSystemLookupFunction.lambda$eval$1(HdpFileSystemLookupFunction.java:59)
> ~[flink-table-blink_2.11-1.12.2-HDP-22.10.09.jar:1.12.2-HDP-22.10.09]
> at
> org.apache.flink.runtime.security.SecurityUtils.runAtOneContext(SecurityUtils.java:100)
> [flink-dist_2.11-1.12.2-HDP-22.10.09.jar:1.12.2-HDP-22.10.09]
> at
> org.apache.flink.table.filesystem.security.HdpFileSystemLookupFunction.eval(HdpFileSystemLookupFunction.java:58)
> [flink-table-blink_2.11-1.12.2-HDP-22.10.09.jar:1.12.2-HDP-22.10.09]
> at LookupFunction$10.flatMap(Unknown Source)
> [flink-table-blink_2.11-1.12.2-HDP-22.10.09.jar:?]
> at
> org.apache.flink.table.runtime.operators.join.lookup.LookupJoinRunner.processElement(LookupJoinRunner.java:81)
> [flink-table-blink_2.11-1.12.2-HDP-22.10.09.jar:1.12.2-HDP-22.10.09]
> at
> org.apache.flink.table.runtime.operators.join.lookup.LookupJoinRunner.processElement(LookupJoinRunner.java:34)
> [flink-table-blink_2.11-1.12.2-HDP-22.10.09.jar:1.12.2-HDP-22.10.09]
> at
> org.apache.flink.streaming.api.operators.ProcessOperator.processElement(ProcessOperator.java:66)
> [flink-dist_2.11-1.12.2-HDP-22.10.09.jar:1.12.2-HDP-22.10.09]
> at
> org.apache.flink.streaming.runtime.tasks.CopyingChainingOutput.pushToOperator(CopyingChainingOutput.java:71)
> [flink-dist_2.11-1.12.2-HDP-22.10.09.jar:1.12.2-HDP-22.10.09]
> at
> org.apache.flink.streaming.runtime.tasks.CopyingChainingOutput.collect(CopyingChainingOutput.java:46)
> [flink-dist_2.11-1.12.2-HDP-22.10.09.jar:1.12.2-HDP-22.10.09]
> at
> org.apache.flink.streaming.runtime.tasks.CopyingChainingOutput.collect(CopyingChainingOutput.java:26)
> [flink-dist_2.11-1.12.2-HDP-22.10.09.jar:1.12.2-HDP-22.10.09]
> at
> org.apache.flink.streaming.api.operators.CountingOutput.collect(CountingOutput.java:50)
> [flink-dist_2.11-1.12.2-HDP-22.10.09.jar:1.12.2-HDP-22.10.09]
> at
> org.apache.flink.streaming.api.operators.CountingOutput.collect(CountingOutput.java:28)
> [flink-dist_2.11-1.12.2-HDP-22.10.09.jar:1.12.2-HDP-22.10.09]
> at StreamExecCalc$7.processElement(Unknown Source)
> [flink-table-blink_2.11-1.12.2-HDP-22.10.09.jar:?]
> at
> org.apache.flink.streaming.runtime.tasks.CopyingChainingOutput.pushToOperator(CopyingChainingOutput.java:71)
> [flink-dist_2.11-1.12.2-HDP-22.10.09.jar:1.12.2-HDP-22.10.09]
> at
> org.apache.flink.streaming.runtime.tasks.CopyingChainingOutput.collect(CopyingChainingOutput.java:46)
> [flink-dist_2.11-1.12.2-HDP-22.10.09.jar:1.12.2-HDP-22.10.09]
> at
> org.apache.flink.streaming.runtime.tasks.CopyingChainingOutput.collect(CopyingChainingOutput.java:26)
> [flink-dist_2.11-1.12.2-HDP-22.10.09.jar:1.12.2-HDP-22.10.09]
> at
> org.apache.flink.streaming.api.operators.CountingOutput.collect(CountingOutput.java:50)
> [flink-dist_2.11-1.12.2-HDP-22.10.09.jar:1.12.2-HDP-22.10.09]
> at
> org.apache.flink.streaming.api.operators.CountingOutput.collect(CountingOutput.java:28)
> [flink-dist_2.11-1.12.2-HDP-22.10.09.jar:1.12.2-HDP-22.10.09]
> at
> org.apache.flink.streaming.api.operators.StreamSourceContexts$ManualWatermarkContext.processAndCollectWithTimestamp(StreamSourceContexts.java:322)
> [flink-dist_2.11-1.12.2-HDP-22.10.09.jar:1.12.2-HDP-22.10.09]
> at
> org.apache.flink.streaming.api.operators.StreamSourceContexts$WatermarkContext.collectWithTimestamp(StreamSourceContexts.java:426)
> [flink-dist_2.11-1.12.2-HDP-22.10.09.jar:1.12.2-HDP-22.10.09]
> at
> org.apache.flink.streaming.connectors.kafka.internals.AbstractFetcher.emitRecordsWithTimestamps(AbstractFetcher.java:365)
> [flink-sql-connector-kafka_2.11-1.12.2-HDP-22.09.20.jar:1.12.2-HDP-22.09.20]
> at
> org.apache.flink.streaming.connectors.kafka.internals.KafkaFetcher.partitionConsumerRecordsHandler(KafkaFetcher.java:183)
> [flink-sql-connector-kafka_2.11-1.12.2-HDP-22.09.20.jar:1.12.2-HDP-22.09.20]
> at
> org.apache.flink.streaming.connectors.kafka.internals.KafkaFetcher.runFetchLoop(KafkaFetcher.java:142)
> [flink-sql-connector-kafka_2.11-1.12.2-HDP-22.09.20.jar:1.12.2-HDP-22.09.20]
> at
> org.apache.flink.streaming.connectors.kafka.FlinkKafkaConsumerBase.run(FlinkKafkaConsumerBase.java:826)
> [flink-sql-connector-kafka_2.11-1.12.2-HDP-22.09.20.jar:1.12.2-HDP-22.09.20]
> at
> org.apache.flink.streaming.api.operators.StreamSource.run(StreamSource.java:110)
> [flink-dist_2.11-1.12.2-HDP-22.10.09.jar:1.12.2-HDP-22.10.09]
> at
> org.apache.flink.streaming.api.operators.StreamSource.run(StreamSource.java:66)
> [flink-dist_2.11-1.12.2-HDP-22.10.09.jar:1.12.2-HDP-22.10.09]
> at
> org.apache.flink.streaming.runtime.tasks.SourceStreamTask$LegacySourceFunctionThread.run(SourceStreamTask.java:263)
> [flink-dist_2.11-1.12.2-HDP-22.10.09.jar:1.12.2-HDP-22.10.09]
> Caused by: java.lang.IllegalArgumentException: Failed to specify server's
> Kerberos principal name
> at
> org.apache.hadoop.security.SaslRpcClient.getServerPrincipal(SaslRpcClient.java:326)
>
> ~[flink-shaded-hadoop-2-uber-2.10.1-HDP-22.09.1-release-10.0.jar:2.10.1-HDP-22.09.1-release-10.0]
> at
> org.apache.hadoop.security.SaslRpcClient.createSaslClient(SaslRpcClient.java:231)
>
> ~[flink-shaded-hadoop-2-uber-2.10.1-HDP-22.09.1-release-10.0.jar:2.10.1-HDP-22.09.1-release-10.0]
> at
> org.apache.hadoop.security.SaslRpcClient.selectSaslClient(SaslRpcClient.java:159)
>
> ~[flink-shaded-hadoop-2-uber-2.10.1-HDP-22.09.1-release-10.0.jar:2.10.1-HDP-22.09.1-release-10.0]
> at
> org.apache.hadoop.security.SaslRpcClient.saslConnect(SaslRpcClient.java:391)
> ~[flink-shaded-hadoop-2-uber-2.10.1-HDP-22.09.1-release-10.0.jar:2.10.1-HDP-22.09.1-release-10.0]
> at
> org.apache.hadoop.ipc.Client$Connection.setupSaslConnection(Client.java:617)
> ~[flink-shaded-hadoop-2-uber-2.10.1-HDP-22.09.1-release-10.0.jar:2.10.1-HDP-22.09.1-release-10.0]
> at org.apache.hadoop.ipc.Client$Connection.access$2200(Client.java:423)
> ~[flink-shaded-hadoop-2-uber-2.10.1-HDP-22.09.1-release-10.0.jar:2.10.1-HDP-22.09.1-release-10.0]
> at org.apache.hadoop.ipc.Client$Connection$2.run(Client.java:829)
> ~[flink-shaded-hadoop-2-uber-2.10.1-HDP-22.09.1-release-10.0.jar:2.10.1-HDP-22.09.1-release-10.0]
> at org.apache.hadoop.ipc.Client$Connection$2.run(Client.java:825)
> ~[flink-shaded-hadoop-2-uber-2.10.1-HDP-22.09.1-release-10.0.jar:2.10.1-HDP-22.09.1-release-10.0]
> at java.security.AccessController.doPrivileged(Native Method)
> ~[?:1.8.0_342]
> at javax.security.auth.Subject.doAs(Subject.java:422) ~[?:1.8.0_342]
> at
> org.apache.hadoop.security.UserGroupInformation.doAs(UserGroupInformation.java:2012)
>
> ~[flink-shaded-hadoop-2-uber-2.10.1-HDP-22.09.1-release-10.0.jar:2.10.1-HDP-22.09.1-release-10.0]
> at
> org.apache.hadoop.ipc.Client$Connection.setupIOstreams(Client.java:825)
> ~[flink-shaded-hadoop-2-uber-2.10.1-HDP-22.09.1-release-10.0.jar:2.10.1-HDP-22.09.1-release-10.0]
> ... 56 more{code}
> The task did not fail, but hive failed to load data into the cache.
> The hadoop version used in the test is 2.10.1. During the test, it was found
> that changing the hadoop version to 2.8.5 did not throw an exception. Only
> after hive data is loaded to the cache for the first time and the task is
> restarted, the restarted task cannot properly load hive data to the cache;
> The test found that if the nth time (n>=2) load the hive data into the cache
> and the task is restarted, the restarted task can normally load the hive data
> into the cache.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
