[HDFS-6312] WebHdfs HA failover is broken on secure clusters

• Type: Bug
• Status: Closed
• Priority: Blocker
• Resolution: Fixed
• Affects Version/s: 2.4.0, 3.0.0-alpha1
• Fix Version/s: 2.5.0
• Component/s: webhdfs
• Target Version/s: 2.5.0
• Hadoop Flags: Reviewed
• Assignee: Daryn Sharp
• Reporter: Daryn Sharp
• Created: 30/Apr/14 15:47
• Updated: 12/May/16 18:18
• Resolved: 19/Jun/14 23:08

Description

When webhdfs does a failover, it blanks out the delegation token. This will cause subsequent operations against the other NN to acquire a new token. Tasks cannot acquire a token (no kerberos credentials) so jobs will fail.

Issue Links

• is duplicated by: HDFS-6510 WebHdfs clients clears the delegation token on retry (for HA), thus failing retry requests