Before I opted for this solution, I made a proof of concept, then the whole
thing was implemented and it was revealed during testing that when
anonymizing
more than ten columns in a table, that does not work at all, and without
any reasonable error message. This is from my point of view not fair from
Oracle side.
Now I have a workaround built in, which cost me two days. But thanks for
the hint of Adric.
Am Mi., 23. Jan. 2019 um 20:10 Uhr schrieb Mark W. Farnham <mwf@xxxxxxxx>:
cool. Learning something new today. I didn’t even know you could obscure
column content with exp/imp.
*From:* oracle-l-bounce@xxxxxxxxxxxxx [mailto:
oracle-l-bounce@xxxxxxxxxxxxx] *On Behalf Of *Mladen Gogala
*Sent:* Wednesday, January 23, 2019 9:43 AM
*To:* Adric Norris; Ahmed
*Cc:* oracle-l@xxxxxxxxxxxxx
*Subject:* Re: Datapump question
With the advent of cloud and DBaaS, data pump becomes awkward. However,
even Oracle 18c maintains the good, old exp/imp executables which can be
used instead. Ttue, data pump versions are much faster, but there is no
need to copy files locally and there are no limits with the number of
columns.
Mladen Gogala
Database Consultant
Tel: (347) 321-1217
On 1/22/2019 3:03:23 PM, Adric Norris <landstander668@xxxxxxxxx> wrote:
I'm not sure what version you're on, but 11.2.0.4 had a limit of something
like 8-10 columns (going from memory, so the exact number is a bit fuzzy).
We opened a SR for it at the time, and Oracle's response was that you need
to purchase Oracle Data Redaction if the built-in Datapump limit is
insufficient.
On Tue, Jan 22, 2019 at 11:47 AM Ahmed <gherrami@xxxxxxxxx> wrote:
I try to export a table using DBMS_DATAPUMP. The table has 20 columns and
I try to anonymise 11 columns using functions from package. Like this way:
dbms_datapump.data_remap(handle => h1, name => 'COLUMN_FUNCTION',
table_name => 'TEST_TABLE', column => 'COL1', function =>
'MYPKG.FU1', schema => USER);
but it seems that work only for 10 columns.
If I try to anonymize more than 10 columns I get ORA-39001 invalid input
argument at datapump job definition.
Is this a bug or maybe a restriction? Any Idea?
Best Regards
Ahmed Fikri
Senior Developer
Stuttgart, Germany
