SNA in Oracle

Hello

When I ask the database $ v Yvert it comes 1.3468E + 13. How can I check the number type TI? I mean how to convert 1.3468E + 13 a numeric value?

Please suggest.

SID3 wrote:

Hello

When I ask the database $ v Yvert it comes 1.3468E + 13. How can I check the number type TI? I mean how to convert 1.3468E + 13 a numeric value?

Please suggest.

Use the function TO_CHAR()

Tags: Database

Similar Questions

How to restore RMAN backups

Hello
I tried to include works of RMAN Oracle 12 c
If I use the following command, I should have the backuppiece and the archivelog saved in a folder:
RMAN > BACKUP database DATA PLUS ARCHIVELOG;
If I ran this command in 4 different days, which I ended up having 4 folders with different inside files
I guess that, in these files, I have the data files and the archivelogs
Now I move these folders to tape, re-cutting the backups and delete expired
Say that in a month, I need one of these backups if I put a band records and I added start with 'c:\path_of _restored_folder _L', so I added all files in the folder that I restored from tape
Now, if I try to restore the basis of all is well but when I run the recover, it is missing some archiving logs it says
If I put all the files for 4 days then he finds the archivelog she needs and pick up without problem
on this basis I understand that in the folder containing a backuppiece of a specific day there is the newspaper of archive that I needed to recover from the DB, is that true? How can I know if the backups folder I need to replace in order to have all the required log archiving?
At the moment my DB retrieves only if I handed out the Ribbon last backup folder, if I handed out the previous, it restores, but is not recover
Also I have this configuration of the oracle recovery 3 days window maintains backups in order to be able to restore and recover the last 3 days and he simply deletes all obsolete if the FRA is full
I tried earlier to set up the window of 3 days as a test recovery
I ran a few backups and after that I ran the 5th FRA leak out of space and the backup fails
I expect rman to remove the old backup and make way for a new that I have 5 backuppiece and the FRA is full, why he not do?
Thank you

The default behavior of RECOVER is to make a full recovery - right until the transactions today. So it must all archivelogs.

If you want to restore and recover to a point in time, use the SET clause to THAT specify a sequence of the newspaper or time or SNA. Oracle will then not get beyond that point in time - that is to say he will not seek further archiving logs.

(Ideally, when you restore an old backup, I also begins to restore the old controlfile, then using this controlfile).

If you use the FRA in this way

a. set db_recovery_file_dest_size and db_recovery_file_dest

b. * not * used a FORMAT clause in the BACKUP command

Oracle will recognize all backups in the FRA (you can check by querying V$ FLASH_RECOVERY_AREA_USAGE before and after the backup).

In this case, Oracle can and will automatically purge outdated backups - that is to say any backup older than the retention period, provided that there is at least a L0 or backup complete closest to the retention period. You should see messages to this effect in the alert of the instance database log file.

Hemant K Collette

System change number
Oracle documentation describes this as

"A stamp that defines a version committed a database to a point in time. Oracle assigns each transaction posted a single SNA. »
or
"When a user commits a transaction, the transaction is assigned a system change number (SCN), which records the Oracle as well as redo the transaction in the redo log entries"

etc.

I'm a bit confused with the definition of the SNA in Oracle and documentation I see personally. I have a test database, locally, which I'm connected to.

When I have a query as follows the V$ DATABASE, it returns the values as

SELECT current_scn IN the database of v$.
1973076

query again after a few seconds and

SELECT current_scn IN the database of v$.
1973095

I'm assuming that the number changes each time a transaction are committed. However, I realize now, the opposite is true... after seeing the above, that I realize that whenever a transaction occurs, it is assigned a SNA, but the change in the value of the CNS is not in the transaction are committed, I guess that it continues to automatically change all seconds. But sometimes it skips several numbers a second too. So I don't know what this change in value is based on?

Can someone explain which depends on the change in value SNA? All links to the Oracle documentation will be also gratefully appreciated!

Concerning

S diallo
sandeshd wrote:
Is continuous commits that cause the RCS to change, or the RCS changes permanently (regardless of the unit of time) and commited attributed to movements can get this YVERT value?

SNA does not increment by itself. This isn't like the constant opportunity tick of a clock. Instead, many functions within the Oracle request "the next SNA.

Every commit request a new SNA. There are also many other changes within the database that also require a new SNA. Each successive application will get the next YVERT derived in a simple increment.

Since the events that require SNA do not occur at regular intervals, you can consider the RCS to the tick of a clock that has an uneven interval between graduations.

However if having a real clock (wall) that records fast enough times (microseconds or even picoseconds, if the processor is fast enough), you can see each YVERT corresponds to a specific time on this real clock. For ordinary clocks without a PS-hand, you might see 5 SNA in a "seconds" SNA and 200 in the next 'second '.

BUT

SNA can be used as a clock.

Perhaps you've seen the clock sometimes recorded on a film. You will enter in the film you see the increase of the clock and you can see events as part of the film linked at this moment. In the same way, you will enter the SNA in the database, you can see changes to database (or 'events') at every moment.

Therefore, to re-create the database at any moment, you reset the database to any YVERT before that moment (perhaps through a backup) and apply you all database changes, YVERT by RCS (perhaps from your journals of recovery) up to the moment where you want to.

RMAN maintains different SNA

Hello
How to backup rman with different SNA?
RMAN backup is a consistent backup unlike the hot backup.
In my case, some data files with different SNA.
Assume that no newspaper archive (removed manually)
If I perform a recovery, how rman can synchronize the different CSS?
Please see HIGN YVERT column

RMAN > backup validate database.

From backup to 26-SEPT.-15

using channel ORA_DISK_1

channel ORA_DISK_1: from complete compressed datafile backup set

channel ORA_DISK_1: specifying datafile (s) in the backup set

..

...

List of data files

=================

State of the file marked corrupt blocks examined empty blocks high YVERT

---- ------ -------------- ------------ --------------- ----------

1 0 13623 218952 60292123 OK

File name: /u02/app/oracle/oradata/orcl/system01.dbf

Type block for lack of treated blocks blocks

---------- -------------- ----------------

Data 175372 0

Index 0 14086

Other 15799 0

State of the file marked corrupt blocks examined empty blocks high YVERT

---- ------ -------------- ------------ --------------- ----------

2 0 21593 126728 60292106 OK

File name: /u02/app/oracle/oradata/orcl/sysaux01.dbf

Type block for lack of treated blocks blocks

---------- -------------- ----------------

Given 41535 0

Index 0 28122

Other 0 35470

State of the file marked corrupt blocks examined empty blocks high YVERT

---- ------ -------------- ------------ --------------- ----------

3 0 19991 963656 60178764 OK

File name: /u01/app/oracle/oradata/orcl/sda01.dbf

Type block for lack of treated blocks blocks

---------- -------------- ----------------

Data 939356 0

Index 0 0

Other 0 4309

State of the file marked corrupt blocks examined empty blocks high YVERT

---- ------ -------------- ------------ --------------- ----------

7 0 6267 6400 37958094 OK

File name: /u02/app/oracle/oradata/orcl/tbs1_01.dbf

Type block for lack of treated blocks blocks

---------- -------------- ----------------

Data 0 0

0 3 index

Other 0 130

State of the file marked corrupt blocks examined empty blocks high YVERT

---- ------ -------------- ------------ --------------- ----------

8 0 6273 6400 37957419 OK

File name: /u02/app/oracle/oradata/orcl/tbs2_01.dbf

Type block for lack of treated blocks blocks

---------- -------------- ----------------

Data 0 0

Index 0 0

Other 0 127

State of the file marked corrupt blocks examined empty blocks high YVERT

---- ------ -------------- ------------ --------------- ----------

9 0 5 6400 60292118 OK

File name: /u02/app/oracle/oradata/orcl/tbs3_01.dbf

Type block for lack of treated blocks blocks

---------- -------------- ----------------

Data 0 5798

Index 0 51

Other 0 546

..

...

Thanks in advance.

(1) why rman retains much difference to SNA in the same backup? (please see light - blue color)

a single file have 60292123

Another has 37958094

A senior CPN is the last CNS or most recent SCN of the data file. When you checked using 'backup validate database', he was 60292123 for the data file 1 and 37958094 for the 7 data file. If you check the same today, it would be different. If there is no activity or transactions (I mean not transaction YVERT here) performed on the respective data files, there was no change in the high SNA. As the dirty blocks are written to the file, the high RCS of the dataifle also continues to change.

In your case, datafile 7 is not used much more when compared with datafile 1 or 2. Just to let you know, the high RCS of any data file is still below the number of change of control point. In the event of a checkpoint, all dirty blocks are written to disk and allocate a checkpoint change number. This change number of control point for all data files remain the same and are updated in the headers of all the data files. A change of checkpoint # is different from that of the CNS high in the data file.

Just to help you understand, here is an excerpt:

The high RCS for datafile 4 a 990377

State of the file marked corrupt blocks examined empty blocks high YVERT

---- ------ -------------- ------------ --------------- ----------

4 0 33 641 990377 OK

File name: /u02/oradata/testdb1/testdb1/users01.dbf

Type block for lack of treated blocks blocks

Created a user with default tablespace users and some changes were made.

SQL > create user identified by oracle Emmanuelle

2 default tablespace users;

Created by the user.

SQL > conn Emmanuelle/oracle

Connected.

SQL > create table test1 (code number);

Table created.

SQL > insert into test1 values (100);

1 line of creation.

SQL > /.

1 line of creation.

SQL > commit;

Validation complete.

Once these changes are written to the file after a post control, high RCS for datafile 4 changed.

State of the file marked corrupt blocks examined empty blocks high YVERT

---- ------ -------------- ------------ --------------- ----------

4 0 25 641 1044005 OK

File name: /u02/oradata/testdb1/testdb1/users01.dbf

Type block for lack of treated blocks blocks

---------- -------------- ----------------

0 20 data

0 2 index

Other 0 593

HTH

-Jonathan Rolland

Oracle 12 c with RMAN database backup

Hi All-
I started back to the top of one of my PDB using utility RMAN following the instructions on the link below.
http://www.Oracle.com/WebFolder/technetwork/tutorials/OBE/DB/12C/R1/PDB/pdb_pitr/pdb_pitr.html
```
--Connect to Oracle RMAN.
rman> Connect target /
--Set the Oracle RMAN backup device type and backup file location.
rman> configure channel device type disk format '/stage/db_backup_files/cdb1/%U';
--Turn on automatic backup of control files.
rman> configure controlfile autobackup on;
--Back up the database and archive logs.
rman>backup database plus archivelog;
rman> exit
```
Back to the top has been done successfully, but when I started with the recovery process he gave me error reporting collection.
Run {set until what SNA = 2263440; restore database connectable to pdb2; retrieve auxiliary destination of the pluggable database pdb2 = "/ stage/db_backup_files/cdb1 '; alter connectable database to pdb2 ;} open resetlogs}
This is the track to RMAN

Recovery Manager: release 12.1.0.2.0 - Production on sat 22 Aug 12:22:44 2015

Copyright (c) 1982, 2014, Oracle and/or its affiliates. All rights reserved.

RMAN > Connect target /.

connected to target database: RELIABLE (DBID = 1028507357)

RMAN > run {}

2 > ALTER DATABASE connectable to MultiPlay NARROW;

3 > set up what SNA = 14899894;

4 > restore database pluggable MultiPlay;

5 > restore database pluggable MultiPlay;

6 > alter database connectable to MultiPlay open resetlogs;

{7 >}

using the control file of the target instead of recovery catalog database

Statement processed

executing command: SET until clause

From restoration to 22 August 15

allocated channel: ORA_DISK_1

channel ORA_DISK_1: SID = 22 type of device = DISK

channel ORA_DISK_1: from the restore backup set data file

channel ORA_DISK_1: specifying datafile (s) to restore from backup set

channel ORA_DISK_1: restore datafile 00035 to E:\APP\ORADATA\RELIABLE\MULTIPLAY\SYSTEM01. DBF

channel ORA_DISK_1: restore datafile 00036 to E:\APP\ORADATA\RELIABLE\MULTIPLAY\SYSAUX01. DBF

channel ORA_DISK_1: restore datafile 00037 to E:\APP\ORADATA\RELIABLE\MULTIPLAY\MULTIPLAY_USERS01. DBF

channel ORA_DISK_1: backup E:\APP\RECOVERY_AREA\RELIABLE\40A433F530464F3BBF2E7C0B1553ECE8\BACKUPSET\2015_08_22\O1_MF_NNNDF_TAG20150 piece reading

822T115328_BXJ5C182_. BKP

channel ORA_DISK_1: piece handle = E:\APP\RECOVERY_AREA\RELIABLE\40A433F530464F3BBF2E7C0B1553ECE8\BACKUPSET\2015_08_22\O1_MF_NNNDF_TAG20150822T115328_BX

J5C182_. Tag BKP = TAG20150822T115328

channel ORA_DISK_1: restored the backup part 1

channel ORA_DISK_1: restore complete, duration: 00:01:05

Restoration finished in 22 August 15

From pick up to August 22, 15

using channel ORA_DISK_1

RMAN-05026: WARNING: following tablespaces set applies to the Point in time specified assuming

List of tablespaces should have UNDO segments

SYSTEM tablespace

Tablespace UNDOTBS1

Creation of automatic instance with SID = "zbnt."

RMAN-00571: ===========================================================

RMAN-00569: = ERROR MESSAGE STACK FOLLOWS =.

RMAN-00571: ===========================================================

RMAN-03002: failure of the command recover at 22/08/2015 12:24:56

ORA-19852: error creating services for the instance auxiliary zbnt (error 0)

ORA-27302: failure occurred:

ORA-27303: additional information: could not start the instance

Thank you
Siradji

Just a wild guess

You do not use this script in a command as Administrator window

Therefore, you won't be able to create any service.

I also have a question: your only troubleshooting activity consisted of hitting CTRL + C and CTRL-V, is not it?

It's so boring to most of the people here do not even TRY to solve their own problems.

-------------

Sybrand Bakker

Senior Oracle DBA

SNA/ora_rowscn issues

Hi all
I was reading AskTom and he talked about the clothing of the SNA at some point. If this is the case, then how great the RCS measure up until it wraps? Where is stored the film count?
On imports, ora_rowscn on the data in the table will get SNA SNA DB and not what has been exported?
The only use of the CPN in a datapump file would be used for flashback_scn? Right?
Thanks for all replies.

All, Woody would have already seen this, but Tim Gorman posted an answer to this same question on Oracle-L, which explains the operation of the base of SNA and rap as well as some interesting facts

https://www.freelists.org/post/Oracle-l/SCNora-rowscn-questions, 2

- -

HTH - Mark D Powell.

SNA on the primary is less than on previous day

Hello guys, talk like that in this forum but there are ways to solve this problem and I need to know if it should be solved or not.
I created the physical standby and it worked very well. created archive_log files had Yvert number exactly as primary, but then my colleague rebooted the virtual machine on which eve has been fixed, so obviously waiting has been closed to the ground.
in the root I used the command: service iptables stop;
then I started the listener in oracle: lsnrctl start;
and then I used this command in sqlplus:
(1)startup nomount;
2)alter database base_de_donnees eve of Mount;
3)ALTER DATABASE RECOVER MANAGED STANDBY DATABASE disconnect from the session;
This bound mode standby and main cause I saw my archive_log on two database generated files BUT:
I knew my last name of file archive_log on primary was 1_4359_870764186.dbf and the dmp Eve last file was 1_4360_870764186.dbf (number of SNA is a times). then I looked more carefully and realized that the last file dbf standby - 1_4360_870764186.dbf probably changed all the time, cause I've refreshed it's change time was always equal to this day., but it did not record cause of changes its size was always the same: 51201 K;

then I tried to change the journal manually and typed on primary: alter system switch logfile; who created on primary 1_4360archive_log_870764186.dbf and Eve 1_4361_870764186.dbf, I looked carefully on standby and realized that it was now 4361 51201 K size and changed every minute, instead of 4360;
then I looked at sizes of 4360 archive_log file on two databases, and they were equal; as was 4359 on databases and all other log files;
so I feel that the log_files are properly reversed, their size and the SNA are exactly the same. but for some reason any backup database creates this last archive_log every minute and there always 51201K size.

in my alert log files is all cool. It is said:
Fri 31 Jul 09:25:35 2015
Archives journal entry 27 additional for each thread 1 4360 rlc 870764186 0xd6315b1a dest ID sequence 2:
RFS [2]: open the newspaper for thread 1 sequence 4361 dbid-701415654 branch of the 870764186
Fri 31 Jul 09:25:39 2015
Recovery Media Log /opt02/archive_log/1_4360_870764186.dbf
Media, recovery waiting for thread 1 sequence 4361 (in transit)
Yes, it says he's waiting for the sequence 4361 which is not created on primary and should not be created in the standby mode, but is created in waiting until this last file, which is of size as I said is 51201K and changes every minute.
(1) the question then is, what is?
(2) why it was caused when I assembled the night before and primary?
(3) it will have a detrimental effect on my database pending? (should I try to fix it, or should I recreate standby database? or should I leave it alone?)

Hello, I think this is the standard behavior, almost in 11g, I checked against 2 separate dataguard, with and without the application of real time, and I got the same result.

which instance standby waiting for the next archivelog, alert, you can see:

Media Recovery wait thread 1 sequence xxxx (in transit)

and you're right, the size, it is not the same as in the primary.

It seems that oracle creates beforehand the archivelog current mode standby with 50 MB and when this archivelog is finalized in the primary and he was transferred to the day before, it is resized to the actual size.

Kind regards.

How know when an Oracle table has been updated last time?

I want to find when the last INSERT, UPDATE, or DELETE statement is executed on a table (for now, in the future, I want to do it in several tables) in an Oracle database.
I created a table and then I updated one of its lines. Now, I have the following code:
SELECT ORA_ROWSCN * FROM test_table;
SELECT SCN_TO_TIMESTAMP (ora_rowscn) in test_table;
Copy the following code, show timestamps of each row, but from the point where they were created. The line I've updated, have the same time other stamps. Why?
Can someone help me please?

Wrong forum!

As the forum title suggests, this forum is for SQL Developer questions and your question has NOTHING to do with this tool.

Please mark this question ANSWER and repost it in the forum Sql and Pl/Sql

SQL and PL/SQL

Before repost you I suggest you read the ORA_ROWSCN Oracle documentation.

http://docs.Oracle.com/CD/B28359_01/server.111/b28286/pseudocolumns007.htm

Copy the following code, show timestamps of each row, but from the point where they were created.

No - this isn't what he does this last link doc treats:

For each line, ORA_ROWSCN returns the number of change of system (SNA) conservative upper limit of the most recent change to the line.

If the value will be updated if a line in a block is changed.

The line I've updated, have the same time other stamps. Why?

The default is to BLOCK, not the level of the line level - there is only ONE value for the block. If a block contains 100 lines, they appear to have the same value, the value for the block.

For tables defined with ROWDEPENDENCIES values are actually stored for each line.

The * views _TAB_MODIFICATIONS will show you the last time a table has changed SINCE the statistics were collected.

https://docs.Oracle.com/CD/B28359_01/server.111/b28320/statviews_2095.htm

If you need more help or explanation post a new thread in the right forum.

Accuracy of Oracle Flashback?

I ran the following script and was surprised to see the SNA does not change only during the 3 commits, I have published. I had the illusion that the RCS would be updated after each commit.
```
drop table alpha cascade constraints;
purge recyclebin;

create table alpha(
cola number,
colb date) enable row movement;

insert into alpha values(1, sysdate);
execute dbms_lock.sleep(30);
commit;
insert into alpha values(2, sysdate);
execute dbms_lock.sleep(30);
commit;
insert into alpha values(3, sysdate);
execute dbms_lock.sleep(30);
commit;
```
When I then tried to do a step backwards to take me where there was only the first line, I actually had to use a lot later at 23:07:55 (see below) either earlier and table brings up no line!
```
db > select ora_rowscn, cola, to_char(colb, 'hh24:mi:ss fmdd-Month-YYYY') from alpha;


ORA_ROWSCN       COLA TO_CHAR(COLB,'HH24:MI:SSFMDD-MONTH-YYYY')
---------- ---------- -----------------------------------------------------
   4939034          1 23:07:55 24-November-2014
   4939034          2 23:08:25 24-November-2014
   4939034          3 23:08:55 24-November-2014
```
```
db > flashback table alpha to timestamp (timestamp '2014-11-24 23:08:30');


Flashback complete.


db > select ora_rowscn, cola, to_char(colb, 'hh24:mi:ss fmdd-Month-YYYY') from alpha;


ORA_ROWSCN       COLA TO_CHAR(COLB,'HH24:MI:SSFMDD-MONTH-YYYY')
---------- ---------- -----------------------------------------------------
   4940017          1 23:07:55 24-November-2014
```
I make the assumption that the RCS is different because of data in the table, return to old data and using a current SNA to do this, but I do not understand why I had to add on a full 35 seconds to come back to the table when the first line had been committed.
I suppose that the return of FLAME is not as accurate as I thought it would be.

It is not quite accurate, as Oracle follows the transaction SNA validated for the block's in the line.

Can you tell us what part is not clear?

If you make a change to ANY line in a block, the RCS is stored for the block.

So if you ask ORA_ROWSCN for ANY line in this block, you will get the SAME NEW value even for the lines that you have NOT changed.

If the table is created using ROWDEPENDENCIES, Oracle uses 6 bytes for each row to store the YVERT value at the level of the line instead of block level. Then if you query the value of a changed row it will be different from the value for the other lines.

You can see that by a simple test in the scott schema using the dept table:

create the table dept_copy rowdependencies in select * from dept

Select deptno, dname, ora_rowscn in dept

DEPTNO, DNAME, ORA_ROWSCN

10, ACCOUNTING, 48004927

20, RESEARCH *, 48004927

30, SALES *, 48004927

40, OPERATIONS, 48004927

Select deptno, dname, ora_rowscn in dept_copy

DEPTNO, DNAME, ORA_ROWSCN

10, ACCOUNTING, 50989490

20, RESEARCH *, 50989490

30, SALES *, 50989490

40, OPERATIONS, 50989490

The value of ORA_ROWSCN is the same for all rows in the DEPT table and the dept_copy table (but a different value of course).

Now, we update a row in each table

update the set dname dept = ' ACCOUNTING * "where DNAME ="ACCOUNTING ".

update set dept_copy = dname "ACCOUNTING *" where DNAME = "ACCOUNTING".

Commit

Select deptno, dname, ora_rowscn in dept

DEPTNO, DNAME, ORA_ROWSCN

10, ACCOUNTING *, 50989863

20, RESEARCH *, 50989863

30, SALES *, 50989863

40, OPERATIONS, 50989863

Select deptno, dname, ora_rowscn in dept_copy

DEPTNO, DNAME, ORA_ROWSCN

10, ACCOUNTING *, 50989863

20, RESEARCH *, 50989490

30, SALES *, 50989490

40, OPERATIONS, 50989490

Hmmm - the DEPT table has a new value for ORA_ROWSCN but it is the same for all lines. This is because it is stored for the block.

But the DEPT_COPY table has a new value for that ONE line. This is because the Oracle caused ROWDEPENDENCIES add a space to store the value for each line separately as if it were another column in the table.

Try this code example and see what you get. Don't forget the VALIDATION once you change the line - the value of the CNS to change until you actually agree the data.

How Controlfile SNA remain synchronized after moving the database with the cold backup

Hello DBAs,
I doubt that:
Every time you pass a cold backup database. We create a new controlfile with new locations of data/redologs etc.
And then we execute "create controlfile reuse...". noresetlogs...; "to create the controlfile.
After that, we will open the database as > > alter database open;
Now the question is:
We are able to open the database without clause "resetlogs.
Here, we create a new controlfile so SNA to controlfile will not be synchronized with the SCN of the datafile and redo the file.
Please let me know how the new controlfiles' SNA is synchronized with the data file and redofile and wea are able to open the database without resetlogs clause.
Oracle: 11.2.0.4
OS: Linux

If your cold backup also includes online redo logs, a RESETLOGS is not necessary.

The controlfile is created if it does not have a SNA, it reads the data file headers the YVERT database. For a cold backup, all the headers of data file are consistent. (This is also the reason why a RECOVERY is not required).

Hemant K Collette

Oracle db host1 and host2 restore failed

Hello
Oracle DB 11 g 2/Centos 6.5 x 64/AANP 7.6
I test the restoration of the BNU 1 host to another host. Although this configuration will as well as the configuration of NBU, I think that this question of my Oracle RMAN.
Configure a database even with the same instance name (the two default installation), the settings for the restoration. Then restore with RMAN.
RMAN > run {allocate channel ch00 type 'sbt_tape';
2 > Send "NB_ORA_CLIENT is host1.lab.tec";.
3 > restore database;
4 > restore database;
5 > output channel ch00;
{6 >}
I have a problem when restoring. After the recovery is complete, I can run and start db OK.
SQL > alter database open resetlogs;
Database altered.
But this is a small test db, I'm worried if I do it in a real environment. Wihout solve this problem, I would like to get troulbe. Hope someone can help with this.
========= logs error =======================================
using the control file of the target instead of recovery catalog database
allocated channel: ch00
channel ch00: SID = 18 device type = SBT_TAPE
channel ch00: Veritas NetBackup for Oracle - version 7.6 (2013111313)
command sent on the channel: ch00
From restoration to 21-SEP-14
From backup implicit intersection to 21-SEP-14
Finished the backup implicit intersection to 21-SEP-14
Starting copy of implicit intersection to 21-SEP-14
Copy of implicit cross ended up at 21-SEP-14
the search for all files in the recovery area
cataloging files...
not cataloged files
channel ch00: from the restore backup set data file
channel ch00: specifying datafile (s) to restore from backup set
channel ch00: restore datafile 00001 to /data/oracle/oradata/tecdb/system01.dbf
channel ch00: restore datafile 00002 to /data/oracle/oradata/tecdb/sysaux01.dbf
channel ch00: restore datafile 00003 to /data/oracle/oradata/tecdb/undotbs01.dbf
channel ch00: restore datafile 00004 in /data/oracle/oradata/tecdb/users01.dbf
channel ch00: restore datafile 00005 to /data/oracle/oradata/tecdb/bigdata01.dbf
channel ch00: reading of the element of backup bk_dTECDB_u07pj38v9_s7_p1_t858891241
channel ch00: piece handle = bk_dTECDB_u07pj38v9_s7_p1_t858891241 tag = TECDB_BKP
channel ch00: restored the backup part 1
channel ch00: restoration complete, duration: 00:01:15
Restoration finished 21-SEP-14
From back to 21-SEP-14
starting media recovery
channel ch00: archived startup log restore to default destination
channel ch00: restoration of archives Journal
Archives journal thread = 1 sequence 9 =
channel ch00: restoration of archives Journal
Archives journal thread = 1 sequence = 10
channel ch00: reading of the element of backup arch_u09pj393g_s9_p1_t858891376
channel ch00: piece handle = arch_u09pj393g_s9_p1_t858891376 tag = TAG20140921T205616
channel ch00: restored the backup part 1
channel ch00: restoration complete, duration: 00:00:25
Archive log file name=/data/oracle/flash_recovery_area/TECDB/archivelog/2014_09_21/o1_mf_1_9_b1y08dhs_.arc thread = 1 sequence 9 =
default channel: archived deletion or newspapers
RECID = STAMP 7 = 858901628 name=/data/oracle/flash_recovery_area/TECDB/archivelog/2014_09_21/o1_mf_1_9_b1y08dhs_.arc archived log file
Archive log file name=/data/oracle/flash_recovery_area/TECDB/archivelog/2014_09_21/o1_mf_1_10_b1y08dmw_.arc thread = 1 sequence = 10
default channel: archived deletion or newspapers
name=/data/oracle/flash_recovery_area/TECDB/archivelog/2014_09_21/o1_mf_1_10_b1y08dmw_.arc RECID archived log file = 6 = 858901628 STAMP
Cannot find the archived log < < < < < < < < < < < < < failed here
Archives journal thread = 1 sequence = 11
output channel: ch00
RMAN-00571: ===========================================================
RMAN-00569: = ERROR MESSAGE STACK FOLLOWS =.
RMAN-00571: ===========================================================
RMAN-03002: failure of the command recover to 21/09/2014 23:47:18
RMAN-06054: unknown request media recovery log archived for thread 1 with sequence 11 and from the 1030445 SNA

If you have not specified a DEFINED point UNTIL Oracle might attempt a COMPLETE recovery. It is possible that at the time of the backup, suite #11 remained a newspaper online redo and not yet archived on. If so, you would SET UNTIL SEQUENCE 11 while Oracle would cease after having applied the #10 sequence

Hemant K Collette

Multiples of transaction with a SNA

Hello
a question:
Executing a statement that performs multiple operations, I'll have a SNA with several Transaction ID.
How is it written for the redolog? I find 'sets' with the same SNA and the different Transaction ID?
In the case of recocery the instruction will be "redone" ordered by RCS and within the CNS by Id of the Transaction?
I found no Docs on that one

Concerning
Christian

Rivay Shamsudeen explains the RCS in http://orainternals.wordpress.com/2012/01/19/scn-what-why-and-how/:
YVERT (System change number) is a primary mechanism for maintaining the consistency of the data in the Oracle database. YVERT is used mainly in the following areas, of course, is not an exhaustive list:
1. Each redo record has a SNA version of do it again in the header to repeat (and recovery records can have non-unique YVERT). Restore records in terms of both threads (as in the case of CARS), recovery will order their order to SNA, essentially maintain a strict sequential order. As explained in my book, each record redo has also several vectors of change.
2. Each data block has also block YVERT (aka version of block). In addition, a vehicle for change in a redo record also provided block YVERT. This means that a vector of change can be applied to one and only one version of the block. Code checks if the YVERT target in a vector of change correspond with the YVERT block before applying redo record. If there is a gap, corruption errors are thrown.
3. Coherence of reading also uses of SNA. Each query is request environment that includes a SNA at the beginning of the query. A session can see transactional changes only if the validation of this SNA transaction is less then the environment request YVERT.
4. Engage. Each commit will generate YVERT, aka commit SCN, which marks a transaction limit. Group validations are also possible.
...

Each change is incremented YVERT?

Not necessarily. Incrementing SNA is not for each change. For example, in the script below, we'll change the table 1000 times, but the generated RCS will be very little.

URGENT: Need to flashback table, how to find SNA for quite some time before the update that's happened?

Hi all
I have a new database 11G, I haven't used before flashback, but my retention time is a day... I have once the developer did a massive update to a table to corrupt.
How can I find the RCS for this time so I can try this table of flashback?
Thank you
Cayenne

Thank you.. I also found that it works:

SQL > select timestamp_to_scn (to_timestamp('20140103152000','YYYYMMDDHH24MISS')) SNA double;

SNA

----------

98344246

Although does NOT work all the time. But that's just what Oracle when you give him a timestamp to flashback to: it converts the timestamp of a SNA and then restores the data based on the SNA.

See table of Flashback in the language SQL doc

http://docs.Oracle.com/CD/B28359_01/server.111/b28286/statements_9012.htm

TIMESTAMP CLAUSE

Specify a value of timestamp corresponding to the point in time to which you want to return to the table. The expr must evaluate to a valid time stamp in the past. The table will flash back to a period of about 3 seconds to the specified timestamp.

This precision "3 seconds" is because of this conversion to SNA - Oracle uses its internal table.

YVERT is accurate, if you know, but you typically do not have.

Read and consider these warnings in app dev guide advanced

http://docs.Oracle.com/CD/B28359_01/AppDev.111/b28424/adfns_flashback.htm

Guidelines for Oracle Flashback Query
- If a potential error of 3 seconds (maximum) is important for Oracle Flashback Query in your application, use a SNA instead of a timestamp. See guidelines on Oracle Flashback technology.
. . .

General guidelines for Oracle Flashback Table

. . .
- For querying the latest data at a specific time, use a SNA. If you use a timestamp, questioned real time can be up to 3 seconds before the time you specify. Oracle database uses the SNA and assigns them to the timestamps to a granularity of 3 seconds.
For example, suppose values SNA 1000 and 1005 are mapped on the timestamps 08:41 and 08:46, respectively. A query for a moment between 08:41 and 08:45:59 is mapped to the SCN 1000; an Oracle Flashback Query at 08:46 is mapped to the SNA 1005.

By reason of this time-to-SNA map, if you specify a time slightly after a DDL (such as a table creation) Oracle database operation can use a SNA which is located just before the DDL operation, causing the error ORA-1466.

. . .

RMAN restore until THE SNA in DATA GUARD environment

Hello
OS = RHL5
DB = 11.2.0.3 primary and 1 physical database paused.
I have a huge request changes tomorrow and its possible that I have an obligation to RESTORE the database 5 hours back because the application.
I think to be adopted following the method of the official documentation of the Oracle (untilClause):

MOUNT FORCE OF START

RUN

{

UNTIL THE SNA 1418901. # The necessary number SNA will be used here.

RESTORE THE DATABASE;

RECOVER THE DATABASE;

}

ALTER DATABASE OPEN RESETLOGS;

I will of course take Level0 backup tonight and LEVEL1 backup before the start of the work.
My humble question is that: If this is the good method I adopt? and if so, then what about physical databases eve? Is that something to restore/recovery too? It is not required or waiting for database necessary to create again (which I'm afraid of)?
Flash back is not enabled.
Pack of thank you for the suggestion.
Best regards

You can enable FLASHBACK DATABASE on major sites and relief and set a RESTORE POINT before the migration of the applications.

If you stop to redo expedition to the France, the waking state is the last SNA that was applied (i.e. 1001). If you need to return the primary SNA 1011 and then resume the delivery of roll forward, Eve can continue to 1002 from totally ignorant. If you flashback primary to a SNA lower, say, 998, you would have to flashback the night before as well (997 or 998) until you can resume redo expedition to the France.

See the Oracle support Note: how Flashback the primary database in standby Configuration (Doc ID 728374.1)

Hemant K Collette

How to detect changes to stop Oracle
Hello

I have a question on how Oracle detects data committed themselves and stop this datafle...

Scenario - 1
-----------------
session 1
------------
Suppose that a Transaction_1 started at 08:30 that contains a select as "SELECT * FROM TABLE" (assuming this query will run for 15 minutes - time to end execution will be 08:45).
session 2
-------------
Then another session, I changed some of the values of the same table (Transaction_1), and I watched the change at 8.35 AM and also comiitted those
changes at 08:40.

As we know each statement is view coherent reading, and each instruction is associated with the number of SNA. Select start of session 1, it will read the data block / who
a number less than SNA number associated with the statement select Yvert.

As the session 2, I modified some data block, so updated the block displays the highest number of SNA SNA number associated with SNA to select session 1
In this case updated block with plu Yvert that select associate SNA session 1, will be read / reconsructed the tablespace undo to give a coherent picture of data to
the selection running on the session 1.

My question is "how coherent reading that happens in the case the other way around? (let me explain in scenario 2)

Scenario 2
--------------

session 1
-------------
at 8:30 I have modified the contents of the table and do not put an end to the transaction (committed or cancelled neigther)

session2
-----------
08:35, I play a select on the table changed on the session 1.

In this session, only committed changes will be picked who did the select query to boot.

My question is, in this case, how oracle detects changes contains who did before SNA of select statement.

Thank you
user9085814 wrote:
Hello

I have a question on how Oracle detects data committed themselves and stop this datafle...

Scenario - 1
-----------------
session 1
------------
Suppose that a Transaction_1 started at 08:30 that contains a select as "SELECT * FROM TABLE" (assuming this query will run for 15 minutes - time to end execution will be 08:45).
session 2
-------------
Then another session, I changed some of the values of the same table (Transaction_1), and I watched the change at 8.35 AM and also comiitted those
changes at 08:40.

As we know each statement is view coherent reading, and each instruction is associated with the number of SNA. Select start of session 1, it will read the data block / who
a number less than SNA number associated with the statement select Yvert.

As the session 2, I modified some data block, so updated the block displays the highest number of SNA SNA number associated with SNA to select session 1
In this case updated block with plu Yvert that select associate SNA session 1, will be read / reconsructed the tablespace undo to give a coherent picture of data to
the selection running on the session 1.

My question is "how coherent reading that happens in the case the other way around? (let me explain in scenario 2)

Scenario 2
--------------

session 1
-------------
at 8:30 I have modified the contents of the table and do not put an end to the transaction (committed or cancelled neigther)

session2
-----------
08:35, I play a select on the table changed on the session 1.

In this session, only committed changes will be picked who did the select query to boot.

My question is, in this case, how oracle detects changes contains who did before SNA of select statement.

The block is checked in correlation with the table of transaction if there of what the status of the transaction. Mind that when you post, the information is the first recorded in the transactions table that the block can be there in the buffer cache or maybe not (checkpointed). If there is no Commit not recorded in the table for the block, it is called uncommitted.

Aman...

SNA in Oracle

Similar Questions

Maybe you are looking for