[FLINK-3271] Using webhdfs in a flink topology throws classnotfound exception

• Type: Bug
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Affects Version/s: 0.10.1
• Fix Version/s: 1.0.0
• Flags: Important
• Assignee: Abhishek Agarwal
• Reporter: Abhishek Agarwal
• Created: 21/Jan/16 18:52
• Updated: 28/Feb/19 13:21
• Resolved: 11/Feb/16 21:02

Description

I was just trying to run a storm topology on flink using flink-storm. I got this exception -

Caused by: java.lang.NoClassDefFoundError: org/mortbay/util/ajax/JSON
	at org.apache.hadoop.hdfs.web.WebHdfsFileSystem.jsonParse(WebHdfsFileSystem.java:325)
	at org.apache.hadoop.hdfs.web.WebHdfsFileSystem$FsPathResponseRunner.getResponse(WebHdfsFileSystem.java:727)
	at org.apache.hadoop.hdfs.web.WebHdfsFileSystem$AbstractRunner.runWithRetry(WebHdfsFileSystem.java:610)
	at org.apache.hadoop.hdfs.web.WebHdfsFileSystem$AbstractRunner.access$100(WebHdfsFileSystem.java:458)
	at org.apache.hadoop.hdfs.web.WebHdfsFileSystem$AbstractRunner$1.run(WebHdfsFileSystem.java:487)
	at java.security.AccessController.doPrivileged(Native Method)
	at javax.security.auth.Subject.doAs(Subject.java:422)
	at org.apache.hadoop.security.UserGroupInformation.doAs(UserGroupInformation.java:1628)
	at org.apache.hadoop.hdfs.web.WebHdfsFileSystem$AbstractRunner.run(WebHdfsFileSystem.java:483)
	at org.apache.hadoop.hdfs.web.WebHdfsFileSystem.listStatus(WebHdfsFileSystem.java:1277)

My topology list some files on hdfs using webhdfs API. org.mortbay.util.ajax.JSON was included in the application uber jar. I noticed that flink loads the application jar in a child classloader. This is what most likely happened -

1. WebHdfsFileSystem class was loaded through parent class loader since it is included in flink-dist.jar.
2. WebHdfsFileSystem has reference to the org.mortbay.util.ajax.JSON but since it is loaded through parent class loader, WebHdfsFileSystem can't read a class in child class loader.

Ideally all the referenced classes should be available in the distribution jar so that these sort of issues may not occur.