Re: CPI3229: Total length of fields in file> 3.5GB ! - Explained.

So the fine folks at IBM 'splained the situation. I'll merely quote their response:


"CPI3229 is a normal informational message when file QADBRSDFR is
created. This file contains the information needed to recreate the
deferred files. DSPFFD shows it contains several fields of data type
BLOB with a field length of 2GB. Together these field lengths total more
than 3.5GB. But I believe these are maximum lengths; typical data will
actually be much less.
Message CPI3229 will always be in the Joblog when the Database System
File QADBRSDFR has to be created or recreated."


Since this restore was being done with a specified DFRID it makes sense that the QADBRSDFR file would need to be created. Indeed the next message in the joblog indicates that file was created. I did not tie the messages together.


- Larry "DrFranken" Bolhuis

www.frankeni.com
www.iDevCloud.com
www.iInTheCloud.com

On 1/10/2013 2:11 PM, Pete Massiello - ML wrote:

Larry,

I have seen this on restores at 6.1, and there wasn't any problems that we found with the file.

Pete

--
Pete Massiello
iTech Solutions
http://www.itechsol.com
http://www.iInTheCloud.com




-----Original Message-----
From: midrange-l-bounces@xxxxxxxxxxxx [mailto:midrange-l-bounces@xxxxxxxxxxxx] On Behalf Of DrFranken
Sent: Thursday, January 10, 2013 12:59 PM
To: Midrange Systems Technical Discussion
Subject: CPI3229: Total length of fields in file> 3.5GB !

IBM i 7.1 on POWER7 with iASP.

Doing multiple library restores to the iASP. In the joblog this message shows up:
============================================
Message ID . . . : CPI3229
Date sent . . . : 01/10/13 Time sent . : 11:29:44

Message . . . . : The total length of the fields of the
file is greater than 3.5 gigabytes.

Cause . . . . . : Input or update is not allowed for a
record that has a total length greater than 3.5 gigabytes.
============================================

There are no objects listed that are not restored. There are no other errors. We believe there are few if any files with record length over
1000 and NONE over 10,000 bytes so a 3.5GB record length is batshit crazy insane.

Our friends Google and the midrange.com archives ring hollow on this error. Has anyone seen this? Honestly at this point I'm not sure there's actually a problem!