[
https://issues.apache.org/jira/browse/NIFI-5557?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16623046#comment-16623046
]
ASF GitHub Bot commented on NIFI-5557:
--------------------------------------
Github user jtstorck commented on a diff in the pull request:
https://github.com/apache/nifi/pull/2971#discussion_r219377632
--- Diff:
nifi-nar-bundles/nifi-hadoop-bundle/nifi-hdfs-processors/src/main/java/org/apache/nifi/processors/hadoop/PutHDFS.java
---
@@ -266,6 +271,16 @@ public Object run() {
throw new
IOException(configuredRootDirPath.toString() + " could not be created");
}
changeOwner(context, hdfs, configuredRootDirPath,
flowFile);
+ } catch (IOException e) {
--- End diff --
Thanks for changing this to use GSSException.getMajor(). I haven't tested
a ticket expiration occurring during the execution of a call to ugi.doAs (as
opposed to the ticket having expired before ugi.doAs is invoked), but I think
it would be a good idea to move this catch block to the top level try/catch
block of the PrivelegedExceptionAction passed to ugi.doAs().
> PutHDFS "GSSException: No valid credentials provided" when krb ticket expires
> -----------------------------------------------------------------------------
>
> Key: NIFI-5557
> URL: https://issues.apache.org/jira/browse/NIFI-5557
> Project: Apache NiFi
> Issue Type: Bug
> Components: Core Framework
> Affects Versions: 1.5.0
> Reporter: Endre Kovacs
> Assignee: Endre Kovacs
> Priority: Major
>
> when using *PutHDFS* processor in a kerberized environment, with a flow
> "traffic" which approximately matches or less frequent then the lifetime of
> the ticket of the principal, we see this in the log:
> {code:java}
> INFO [Timer-Driven Process Thread-4] o.a.h.io.retry.RetryInvocationHandler
> Exception while invoking getFileInfo of class
> ClientNamenodeProtocolTranslatorPB over host2/ip2:8020 after 13 fail over
> attempts. Trying to fail over immediately.
> java.io.IOException: Failed on local exception: java.io.IOException: Couldn't
> setup connection for princi...@example.com to host2.example.com/ip2:8020;
> Host Details : local host is: "host1.example.com/ip1"; destination host is:
> "host2.example.com":8020;
> at org.apache.hadoop.net.NetUtils.wrapException(NetUtils.java:776)
> at org.apache.hadoop.ipc.Client.call(Client.java:1479)
> at org.apache.hadoop.ipc.Client.call(Client.java:1412)
> at
> org.apache.hadoop.ipc.ProtobufRpcEngine$Invoker.invoke(ProtobufRpcEngine.java:229)
> at com.sun.proxy.$Proxy134.getFileInfo(Unknown Source)
> at
> org.apache.hadoop.hdfs.protocolPB.ClientNamenodeProtocolTranslatorPB.getFileInfo(ClientNamenodeProtocolTranslatorPB.java:771)
> at sun.reflect.GeneratedMethodAccessor344.invoke(Unknown Source)
> at
> sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
> at java.lang.reflect.Method.invoke(Method.java:498)
> at
> org.apache.hadoop.io.retry.RetryInvocationHandler.invokeMethod(RetryInvocationHandler.java:191)
> at
> org.apache.hadoop.io.retry.RetryInvocationHandler.invoke(RetryInvocationHandler.java:102)
> at com.sun.proxy.$Proxy135.getFileInfo(Unknown Source)
> at org.apache.hadoop.hdfs.DFSClient.getFileInfo(DFSClient.java:2108)
> at
> org.apache.hadoop.hdfs.DistributedFileSystem$22.doCall(DistributedFileSystem.java:1305)
> at
> org.apache.hadoop.hdfs.DistributedFileSystem$22.doCall(DistributedFileSystem.java:1301)
> at
> org.apache.hadoop.fs.FileSystemLinkResolver.resolve(FileSystemLinkResolver.java:81)
> at
> org.apache.hadoop.hdfs.DistributedFileSystem.getFileStatus(DistributedFileSystem.java:1317)
> at org.apache.nifi.processors.hadoop.PutHDFS$1.run(PutHDFS.java:254)
> at java.security.AccessController.doPrivileged(Native Method)
> at javax.security.auth.Subject.doAs(Subject.java:360)
> at
> org.apache.hadoop.security.UserGroupInformation.doAs(UserGroupInformation.java:1678)
> at org.apache.nifi.processors.hadoop.PutHDFS.onTrigger(PutHDFS.java:222)
> {code}
> and the flowfile is routed to failure relationship.
> *To reproduce:*
> Create a principal in your KDC with two minutes ticket lifetime,
> and set up a similar flow:
> {code:java}
> GetFile => putHDFS ----- success----- -> logAttributes
> \
> fail
> \
> -> logAttributes
> {code}
> copy a file to the input directory of the getFile processor. If the influx
> of the flowfile is much more frequent, then the expiration time of the ticket:
> {code:java}
> watch -n 5 "cp book.txt /path/to/input"
> {code}
> then the flow will successfully run without issue.
> If we adjust this, to:
> {code:java}
> watch -n 121 "cp book.txt /path/to/input"
> {code}
> then we will observe this issue.
>
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
