1

Topic: Whether there will be an inaccessibility at a forehead of segments in another the TOP?

Good afternoon!
Oracle SE 11g
There is a task:  LOB a segment in new tabular space for the specified table.
The decision is selected:
ALTER TABLE TEST MOVE LOB (TEST_NAME) STORE AS (TABLESPACE EXAMPLE);
alter index SCHEMA.INDEX_NAME_PK rebuild tablespace TS_LOB_NEW;
Question: whether there will be an inaccessibility at LOB-segment relocation? ?
The size of a LOB-segment about 100 Gb.

2

Re: Whether there will be an inaccessibility at a forehead of segments in another the TOP?

Lobber;
It is a little strange that the decision at first is selected, and such questions are then asked. Though with SE the choice in general is not present.
What does availability mean? On reading availability will be, same Oracle.
As a variant (if the code of applications allows), it is possible to get in the table the second LOB-field in necessary , slowly to copy in it the data from the first, to delete the first and to rename the second into the first.

3

Re: Whether there will be an inaccessibility at a forehead of segments in another the TOP?

I try to carry out transfer of lob-segments to other HARDWARE.
I receive the message that picture UNDO became outdated.
Undo Retention costs 7200, Retention Guarantee it is used, I expand UNDO,
I increase Retention. The problem remains

4

Re: Whether there will be an inaccessibility at a forehead of segments in another the TOP?

lobber wrote:

Undo Retention costs 7200, Retention Guarantee it is used, I expand UNDO,
I increase Retention. The problem remains

And to dock to smoke not comme il faut? Out-of-row LOB to sneeze oe on UNDO and its size - out-of-row LOB UNDO always in ao LOB ceee. A question: you change only undo_retention or undo_retention + LOB retention? If only undo_retention that LOB it on a drum. If you want e LOB retention that you change undo_retention plus:

alter table tbl modify lob (lob_column) (retention);

SY.

5

Re: Whether there will be an inaccessibility at a forehead of segments in another the TOP?

Checked a DB on beaten units by means of RMAN backup validate database;
Beaten units it was revealed not. At the same time systematically are
Beaten LOB/CLOB in a DB. The question dares simply: the beaten picture is deleted or
The document from basis. But permanently it would not be desirable to do it. There is a suspicion that business
In beaten sectors on a disk. A DB on  VMWare. There is an access only to the virtual
To the machine and all. It is the industrial environment.
Prompt, what method it is better to check up disks, without unmount?
Or to make so that idle time was minimum?
By the way, error Snapshot Too Old has been connected with beaten LOB/CLOB.

6

Re: Whether there will be an inaccessibility at a forehead of segments in another the TOP?

BigBudda wrote:

By the way, error Snapshot Too Old has been connected with beaten LOB/CLOB.

as interesting that! smile

7

Re: Whether there will be an inaccessibility at a forehead of segments in another the TOP?

Q.Tarantino;
Still as it is interesting, especially after you will read about UNDO and Retention, and you think that a problem
snapshot too old because of UNDO, instead of a beaten picture in a DB.

8

Re: Whether there will be an inaccessibility at a forehead of segments in another the TOP?

Prompt, if check of disks showed that beaten sectors are not present, with disks all ,
And in a DB there are no beaten units, whether there can be other reasons which lead beaten LOB/CLOB?
Whether incorrect operation of application in a DB can save beaten pictures and a context?
Someone can already faced the similar? How solved a problem?

9

Re: Whether there will be an inaccessibility at a forehead of segments in another the TOP?

BigBudda wrote:

Still as it is interesting, especially after you will read about UNDO and Retention, and you think that a problem
snapshot too old because of UNDO, instead of a beaten picture in a DB.

Continue to anneal.

BigBudda wrote:

and in a DB there are no beaten units, whether there can be other reasons which lead beaten LOB/CLOB?

And how you revealed that LOB appeared beaten?

10

Re: Whether there will be an inaccessibility at a forehead of segments in another the TOP?

wrote:

and how you revealed that LOB appeared beaten?

So:

create table corrupt_lobs (corrupt_rowid rowid);
declare
error_1578 exception;
error_1555 exception;
error_22922 exception;
pragma exception_init (error_1578,-1578);
pragma exception_init (error_1555,-1555);
pragma exception_init (error_22922,-22922);
n number;
begin
for cursor_lob in (select rowid r, document from LOBDATA) loop
begin
n: = dbms_lob.instr (cursor_lob.document, hextoraw (' 889911 '));
exception
when error_1578 then
insert into corrupt_lobs values (cursor_lob.r);
commit;
when error_1555 then
insert into corrupt_lobs values (cursor_lob.r);
commit;
when error_22922 then
insert into corrupt_lobs values (cursor_lob.r);
commit;
end;
end loop;
end;
/
update LOBDATA set document = empty_blob ()
where rowid in (select corrupt_lobs from corrupt_lobs);

http://oradmin.blogspot.ru/2009/08/ora- … lback.html

11

Re: Whether there will be an inaccessibility at a forehead of segments in another the TOP?

BigBudda;
And what version  that?

12

Re: Whether there will be an inaccessibility at a forehead of segments in another the TOP?

wrote:

and what version  that?

11.2.0.4

13

Re: Whether there will be an inaccessibility at a forehead of segments in another the TOP?

BigBudda wrote:

it is passed...
11.2.0.4

Strange. The last time met that   without the visible reasons of the version so on 8th...
On a metalink did not try to look?