After some round of testing came to know following thing.
We created the single principle for both the nodes, then restarted the cluster then we did not seen checksum failure exceptions and both nodes are working fine..
Might be we need to create single principle for both namenodes..
From: Subroto [[EMAIL PROTECTED]]
Sent: Wednesday, May 08, 2013 3:30 PM
To: [EMAIL PROTECTED]
Subject: Re: Kerboes--Checksum failed----HDFS IN HA mode
Not very sure about the problem but, the following link can show some light:
On May 8, 2013, at 6:43 AM, Brahma Reddy Battula wrote:
Caused by: KrbException: Checksum failed