Joe,

I can reproduce this easily.  Set up a connection to a kerberized Hive 
instance.  After 24 hours you will get errors about an expired TGT.  Restarting 
the Ni-Fi process is the only way I've found to get it to renew the TGT.

Jonathan Schneider
Hadoop/UNIX Administrator, STSC
SCL Health
17501 W. 98th St, Pillars 25-33
Lenexa, KS  66219
P: 913.895.2999
[email protected] 
www.sclhealthsystem.org
 



-----Original Message-----
From: Joe Witt [mailto:[email protected]] 
Sent: Wednesday, January 10, 2018 9:53 AM
To: [email protected]
Subject: [EXTERNAL EMAIL]Re: Kerberos hive failure to renew tickets

*** CAUTION!  This email came from outside SCL Health. Do not open attachments 
or click links if you do not recognize the sender. ***

Georg

We'd need to see what you mean to really understand.  Can you please share NiFi 
logs directory content and if the flow is stuck/locked up please share a nifi 
thread dump which will be in the logs if you first run bin/nifi.sh dump.

thanks

On Wed, Jan 10, 2018 at 8:50 AM, Georg Heiler <[email protected]> wrote:
> Hi
> In production I observe problems with ticket renewal for the nifi hive 
> processor.
>
> A workaround is to restart the hive service but that doesn't seem right.
>
> Is there a real fix for this problem?
>
> Best Georg

Reply via email to