Veritas-bu

[Veritas-bu] Tracking Etrack Incident numbers

2006-10-27 08:43:22
Subject: [Veritas-bu] Tracking Etrack Incident numbers
From: dave_norrie at agilent.com (dave_norrie AT agilent DOT com)
Date: Fri, 27 Oct 2006 13:43:22 +0100
Hi All, 

 

 I raised a call to investigate an exit status 84 error and we resolved it by 
using the workaround detailed below. Problem is that it's totally killing the 
performance of a bpduplicate session, which results in me being unable to 
duplicate a full backup over the weekend. Does anyone know of a way to be 
informed (other than reading the next MP release notes) when a more permanent 
fix is put in place? I'm assuming that the workaround is only a temporary 
solution. Here's the full Etrack Incident number and details:-

 

 
--------------------------------------------------------------------------------

Etrack Incident = ET646985

 

Description:

    NDMP backups failed with a status 84 when writing a TIR fragment.

 

    The bptm log showed an error where the expected block position is one off.

    This error occurred in the function write_data_tir().

 

Workaround:

    To avoid this issue, add the following line to the file

    <installdir>/netbackup/db/config/ndmp.cfg

 

        NDMP_MOVER_CLIENT_DISABLE

 

    This will have a minimal impact on backups. However it will cause a

    significant performance impact to dup/verify/import when using

    NDMP drives.

--------------------------------------------------------------------------------

 

 Cheers,

 

Dave

 

 

 

-- 
Dave Norrie              One IT/Technical Computing Services
Agilent Technologies UK Ltd. 
South Queensferry, Scotland 
Telnet 313-3277        Telephone +44 (0) 131 331 6277

 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: 
http://mailman.eng.auburn.edu/pipermail/veritas-bu/attachments/20061027/89444bb2/attachment-0001.html

<Prev in Thread] Current Thread [Next in Thread>
  • [Veritas-bu] Tracking Etrack Incident numbers, dave_norrie AT agilent DOT com <=