Re: Oracle EBS fnd_file problem
Date: Thu, 28 Feb 2013 13:48:18 -0500
Message-Id: <512F60220200000B004FD073_at_groupwise.gcrta.org>
1, failing job is custom. Others shown by the SQL are the seeded workflow background process. 2. custom job always fails
3. very old cloned environment
>>> John Piwowar <jpiwowar_at_gmail.com> 2/28/13 1:30 PM >>>
Hi Jeff,
Seems a bit odd to me that the multiple requests would be referencing the same temp file name, but I don't know everything. :)
Couple of things to consider:
1) Are the requests that write to this file from custom code, or is the "workflow process" a seeded concurrent request? 2) What's the exit status of the concurrent requests trying to access the file? Are they all failing, or some succeeding? If the latter, it might be that multiple requests are trying to write to the same temp file simultaneously, which could result in the sort of errors you're seeing. 3) You don't mention whether this is production or a cloned environment (though the 'orap' instance name makes me suspect prod :). Any chance this temp file is an artifact from a previous clone, or you have a cloned env pointing its APPLPTMP to the wrong directory and it's in contention with orap?
Regards,
John P.
On Thu, Feb 28, 2013 at 7:30 AM, Jeffrey Beckstrom <JBECKSTROM_at_gcrta.org> wrote:
I looked at a log from a run of this job on 2/19 and it is referencing the SAME tmp file as in the error message from a 2/11 log! Is this supposed to happen this way?
I did some googling and found a script which I altered to do:
select cp.plsql_log, cr.request_id
from apps_applmgr.fnd_concurrent_requests cr, apps_applmgr.fnd_concurrent_processes cp
where cp.plsql_log is not null
and cp.concurrent_process_id = cr.controlling_manager
and cp.plsql_log = 'l0127129.tmp'
order by cr.request_id;
This script returned a lot of rows.
-- http://www.freelists.org/webpage/oracle-lReceived on Thu Feb 28 2013 - 19:48:18 CET