AW: small bug related to deferred_segment_creation parameter
Date: Thu, 22 Jul 2021 22:12:53 +0200 (CEST)
Message-ID: <1626984772994.4188548.b51e9cadf6d0d14ac327db671e396eb31193c71b_at_spica.telekom.de>
I managed to start my RAC 19.0.0. The bug is still there.
Unfortunately I don't have any 19.3 instance. But I don't think that Oracle
would solve the problem there. Maybe it is considered by oracle as feature
and not as bug.
Thanks and best regards
-----Original-Nachricht-----
have just test it in 12.2.0.1 --> the bug is still there.
I'm not sure if Oracle or another user has even noticed this problem to be
solved in 19.3. (Who cares in times of cheap HDDs).
I have a 19.3 (a RAC one), but I'm struggling to get it up (since months I
didn't use it). Maybe someone can check whether the bug is still in 19.3
Best regards
-----Original-Nachricht-----
I'm not surprised that you didn't get a response in less than 3 hours -
especially since you were talking about 11.2.0.4 and 12.1.0.2.
However, it may be a little more interesting to point out that this example
does the same thing in 19.11.0.0 - and still creates the segments even if
you include the "segment creation deferred" option with each of the listed
partitions in the split.
Regards
On Thu, 22 Jul 2021 at 12:53, ahmed.fikri_at_t-online.de
it seems that the deferred_segment_creation parameter does not work
properly when splitting a table partition that contains a clob field.
If the partition is empty and has no segment, splitting it creates two
segments for the old and new partitions.
I tested this using this block in 11.2.0.4 and 12.1.0.2 (also attached
and run with a non dba user):
DECLARE
END;
Is this already known? Any Idea?
Best regards
Ahmed
Betreff: AW: small bug related to deferred_segment_creation parameter
Datum: 2021-07-22T21:44:50+0200
Von: "ahmed.fikri_at_t-online.de" <ahmed.fikri_at_t-online.de>
An: "Jonathan Lewis" <jlewisoracle_at_gmail.com>, "list, oracle"
<oracle-l_at_freelists.org>
Ahmed
Betreff: AW: small bug related to deferred_segment_creation parameter
Datum: 2021-07-22T21:23:14+0200
Von: "ahmed.fikri_at_t-online.de" <ahmed.fikri_at_t-online.de>
An: "Jonathan Lewis" <jlewisoracle_at_gmail.com>, "list, oracle"
<oracle-l_at_freelists.org>
Jonathan Lewis
<mailto:ahmed.fikri_at_t-online.de> <ahmed.fikri_at_t-online.de
<mailto:ahmed.fikri_at_t-online.de> > wrote:
Hi everyone,
PROCEDURE ex(cmd VARCHAR2) IS BEGIN dbms_output.put_line(cmd); EXECUTE
IMMEDIATE cmd; END;
PROCEDURE ex_ignore_error(cmd VARCHAR2) IS BEGIN ex(cmd); EXCEPTION
WHEN OTHERS THEN NULL; END;
PROCEDURE split_part(tbl VARCHAR2,p_part_val VARCHAR2) IS BEGIN
ex('alter table '||tbl||' split partition maxvalue at
(to_date('''||p_part_val||''',''yyyymmdd''))
into (partition part_'||p_part_val||' tablespace users, partition
maxvalue tablespace users)') ;END;
BEGIN
ex('alter session set deferred_segment_creation = true');
ex_ignore_error('drop table tbl_test purge');
ex_ignore_error('drop table tbl_test2 purge');
/* create range-partitioned table without clob field */
ex('create table tbl_test(id number, part_key date) tablespace users
partition by range(part_key) (partition maxvalue values less than
(maxvalue))');
/* create range-partitoned table having a clob field */
ex('create table tbl_test2(id number, part_key date, clob_col clob)
tablespace users partition by range(part_key) (partition maxvalue values
less than (maxvalue))');
/* the table is empty. Splitting the maxvalue partition for the table
without clob */
split_part('tbl_test','20200101');
/* until here no segments is generated for tbl_test and tbl_test2 tables
spltting the maxvalue partition for the table having clob field */
split_part('tbl_test2','20200101');
For me, this behavior is annoying and is considered a bug.
Ahmed
--
http://www.freelists.org/webpage/oracle-l
Received on Thu Jul 22 2021 - 22:12:53 CEST