[HDFS-16503] Should verify whether the path name is valid in the WebHDFS

• Type: Bug
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Affects Version/s: 3.4.0
• Fix Version/s: 3.4.0
• Component/s: webhdfs
• Labels: pull-request-available
• Target Version/s: 3.4.0
• Hadoop Flags: Reviewed
• Assignee: Tao Li
• Reporter: Tao Li
• Created: 14/Mar/22 01:39
• Updated: 11/Feb/24 23:35
• Resolved: 21/Mar/22 07:11

Description

When creating a file using WebHDFS, there are two main steps:
1. Obtain the location of the Datanode to be written.
2. Put the file to this location.

Currently NameNodeRpcServer verifies that pathName is valid, but NamenodeWebHdfsMethods and RouterWebHdfsMethods do not.

So if we use an invalid path(such as duplicated slash), the first step returns success, but the second step throws an InvalidPathException. IMO, we should also do the validation in WebHdfs, which is consistent with the NameNodeRpcServer.

The same webHDFS operations are: CREATE, APPEND, OPEN, GETFILECHECKSUM. So we can add DFSUtil.isValidName to redirectURI for NamenodeWebHdfsMethods and RouterWebHdfsMethods.

Issue Links

• links to: GitHub Pull Request #4067