AnsweredAssumed Answered

Export/Import process anomaly

Question asked by coherentkris on Jun 18, 2015
Latest reply on Jun 19, 2015 by coherentkris

I am trying to diagnose an issue with a stable data transfer script that recently reported inconsistencies.

 

Process goal is to move data from ESS shadow tables (oracle) into FM tables.

The process is fired nightly via server side script and has been running perfectly for months.

Their are 12 tables where data is moved one table at a time.

Two weeks ago one and only one of 12 table data sync routines showed problems.

The destination table is NOT nulled before import.

 

In a nutshell:

Go to source (ESS) layout

Capture the count of records (10516)

Export all of the data into a .mer file

Record the export last error code (0)

Go to destination layout (FM table)

Import matching on three fields, auto enter calcs on, add remaining records as new

Record the import last error code (0)

Count records in destination (10515)

Compare destination record count with source record count.

 

Their is no garbage ( ascii control characters ) in the source data.

The source data contains no empty values in any of the 3 match fields.

The 3 match fields form a valid composite key in the source table (no duplicates exist).

Validated that the .mer file contains the correct record count (10516).

Import dialog character set is Windows (ANSI).

Output dialog character set is Windows (ANSI) and no formatting is retained.

import last error code = 0.

export last error code = 0.

Their is no evidence of corruption (server did not experience any unplanned shutdowns, consistency check of the file shows no issues, recover on the file shows no significant issues)

 

After run:

Source table total record count is 10516.

Destination table total record count is 10515.

These numbers should match.

 

Unfortunately i am unable to post the file or the script.

 

Any thoughts?

 

KM

 

Message was edited by: Kris Musshorn The 3 match fields form a valid composite key in the source table (no duplicates exist).

Outcomes