How ArchiveLog mode impacts performance?

How much it negate the performance in nonarchivelog mode?

Providing your archived logfiles are on a separate disk from the Oracle datafiles, I don't think the performance impact is anything to worry about. A separate ARCH process deals with the actual archiving process. You may want to check that the number of log groups you have is sufficient in relation to the transaction volume of your database. For example, if you have 3 logfile groups (#1, #2 and #3) and log #1 is still being archived, then when log #3 fills, your database will hang until log #1 is available for use again.
Still, just consider the performance hit of only being able to recover the DB to the last cold backup... ;-)
null

Similar Messages

How to use OEM grid control to find any database not in archivelog mode?

Hello,
currently we have 130 databases, say, if I want to pick out any database currently NOT in archivelog mode, is there a quick way to find out ?
by using OEM grid control
or
some trick similar,
but really don't need to check each database manually.
thank you very much!

This query works in 12c - please run as SYSMAN.
It is based on undocumented assumptions though +( AND key_value = 'NOARCHIVELOG')+.
SELECT *
FROM SYSMAN.MGMT$GROUP_DERIVED_MEMBERSHIPS O ,
    SYSMAN.MGMT$TARGET T ,
    sysman.MGMT$AVAILABILITY_CURRENT st
WHERE
o.member_target_type     IN ('oracle_database', 'rac_database')
AND ( t.target_type           ='rac_database'
OR (t.target_type             ='oracle_database'
AND t.type_qualifier3        != 'RACINST'))
AND o. member_target_guid     = t.target_guid
AND t.target_guid             = st.target_guid
AND st.availability_status    = 'Target Up'
AND ( EXISTS
    (SELECT *
    FROM sysman.mgmt$metric_current i
    WHERE I.TARGET_GUID = O.MEMBER_TARGET_GUID
    AND metric_name     = 'archFull'
    AND Metric_Column   = 'archTotal'
    AND metric_label    = 'Archive Area'
    AND column_label    = 'Total Archive Area (KB)'
    AND key_value       = 'NOARCHIVELOG'
    ));Another way to achieve the goal is to push extended metrics (SELECT LOG_MODE FROM V$DATABASE) on each DB, and monitor the values they return to the repository. This is an example for a (slightly) different problem - http://iiotzov.wordpress.com/2012/08/08/how-to-use-the-new-oem-12c-metric-extensions-to-enforce-enterprise-wide-custom-policies .
OEM Repository has many other great uses as well - http://iiotzov.files.wordpress.com/2012/05/oem-repository-a-second-look.doc , http://iiotzov.files.wordpress.com/2011/08/iotzov_oem_repository.pdf
Iordan Iotzov
http://iiotzov.wordpress.com/

Performance deteriorates with archivelog mode

hi,
i want to implement an incremental backup procedure for a 9i database, but the database is in noarchivelog mode.
switching to archivelog mode results in a considerable loss of performance.
is there anything that can be done to either make incremental backups in it's current state, or speed up performance once it is in archivelog mode?
thanks
santosh sewlal

You will be able to get a lot of information about all books in this link :
http://otn.oracle.com/pls/db92/db92.docindex?remark=homepage
[email protected]
Joel P�rez

Performance hit when running in ARCHIVELOG mode.

What is the performance hit when running in ARCHIVELOG mode?
Thank you,
David

I am not one to disagree with Tom Kyte (unless I think he is wrong :) ), and I am not going to disagree here. I do caution the simplistic answer that the hit is negligible. I commend the respondent who qualified that answer with a discussion of I/O.
I have come across more than one situation where archive logging was a performance hit because of the associated I/O. Many want to put archive logs on cheaper storage and do not recognize that not only can this slow a system but that it could become a major issue resulting in a system that hangs until the logs are written. A better solution for these folks is to write to fast storage and have a secondary process that offloads those logs to the slower storage.
Let us also not assume that the archive location is local disk. It might be that an archive location is remote, such as with log shipping or NFS. Network latency can become an issue.
There are many things to consider as there always are. I suppose with any answer, even if simple, one could spin it with some obscure situation that makes the simple answer inappropriate. Having seen some burned by this issue, I chose to elaborate, and I appreciate your indulgence.
Chris

Backup in ARCHIVELOG Mode

Dear Sirs,
Actually I've a non archivelog mode RDBMS, but I wan't to make a
hot backup. Could I make a hot backup after I turn the
ARCHIVELOG mode ON, then turn it off again after Back - Up ?
And how do I make the hot backup ?
Thank you for the attention and help.
Regards,
Tonny

You could turn archivelog mode on, do the backup, and then turn
archivelog mode off. This really defeats the purpose of hot
backups though. You would have to know the SCN, archivelog
file, or time to do a point-in-time recovery manually, or if you
are using RMAN knowing the time will help. I think RMAN will
recover until the last archivelog if you don't specify a time or
SCN. Keep in mind that you will lose all the changes since the
last backup if you turn archivelog mode off after the backup and
your database crashes inbetween the next scheduled backup. If
you are going through the trouble for turning archivelog mode on
and then off I would perform a cold backup. You have to
shutdown the database with normal or immediate, mount the
database, turn archvelog mode on/off and then open the
database. So why not just do a cold backup?

Archivelog mode during upgrade

HI All,
During upgrading the database from 10g to 11g... I will put my database from noarchivelog mode to archivelog mode to save the disk space.
Does this task improves or speedsup the process of upgradation or any impact.
Suggestions please.
regards,
TEJAS

when DB is archivelog mode and you put DB in startup upgrade mode.. and execute catupgrd.sql script it starts updating dictionary objects and generates archivelogs ..
Hence i put the DB in noarchive log mode..
Any performance impact by putting the upgrade process in noarchivelog mode in terms of time duration the backup is performed.No performance impact, But recommended to put in Archivelog mode, even though if you have good backup.
Also apart from this thread, have a question, The catupgrd.sql script execution duration depends upon the data disctionary contents right not on the DB size ? i am having this question. plz answer.The upgrade script creates and alters certain data dictionary tables. It also upgrades or installs the following database components in the new Oracle Database 11g Release 2 (11.2) database:
Source:-
Answer is here in detail.. http://docs.oracle.com/cd/E11882_01/server.112/e10819/upgrade.htm#sthref184

Archivelog mode Configuration

Hi
I am using Oracle10g , and OS is Windows XP. i had tried to enable it to archive mode by using the following
SQL> SELECT LOG_MODE FROM SYS.V$DATABASE;
LOG_MODE
NOARCHIVELOG
Then by adding the following lines in the Pfile
$ORACLE_HOME/admin/SID/pfile. I'll add the following lines to the end of the file:
# Archive Log Destinations -benr(10/15/04)
log_archive_dest_1='location=/u02/oradata/cuddle/archive'
log_archive_start=TRUE
startup the database in mount mode and put it in archivelog mode
SQL> startup mount
SQL> alter database archivelog;
Database altered.
SQL> alter database open;
Database altered.
SQL> select log_mode from v$database;
LOG_MODE
ARCHIVELOG
when i log in to the db, i only can log in as sys user.when log in as sys user i got this error :
"ORA-00257: archiver error. Connect internal only,until freed."
What's the possible problem? And how to solve it?
Regards,
[email protected]

hi surya,
i think your are not using your pfile during startup, and the changes you have made to your pfile does not take effect.
the parameter log_archive_start=TRUE is no longer supported in 10g.you should get a notification of parameter deprecated during startup.
you issued the command : startup , so your pfile is not being used.
you should issue : archive log list to see your archive destination
if it is default, it should be db_recovery_file_dest
next issue show parameter db_recovery_file_dest
check the space availability of the directory by issuing df -h, df -k
you can manually removed your old archive log but after you need to perform crosscheck from rman.
best practise is to do it using rman during your scheduled backup.
br,
mrak

Problem running oracle 9i database in archivelog mode - database freezes

Greetings,
I recently needed to change a (test) database to archivelog mode to make online back-ups of this database. This posed no real problem although I seem to be unable to set the database to automatic archiving eventhough I edited the init.ora file (translation of the "archive log list" output between square brackets):
SQL> archive log list
Databaselogmodus Archiveermodus [Databaselogmode: Archivemode]
Automatisch archiveren Gedeactiveerd [Automatic archiving: Deactivated]
Archiveerbestemming O:\oracle\ora92\RDBMS [Archivedestionation: O:\oracle\ora92\RDBMS]
Oudste on-line logreeks 12306
Volgende logreeks voor archivering 12306
Huidige logreeks 12308
I added this to my init.ora file:
# Archvie Log Destinations
log_archive_start=TRUE
log_archive_dest_1=%ORACLE_HOME%\database\archive
log_archive_format="%%ORACLE_SID%%t_%s.arc"
However, this is not the real problem (maybe it has something to do with it, which is why I mention it aswell), the real problem is that my database freezes when the archive log file is "full". When the archive file reaches ~102.400Kb it just freezes up and I need to restart the database.
The following error is written to the alert_sid.log file:
Thu Apr 17 09:19:55 2008
Thread 1 cannot allocate new log, sequence 12309
All online logs needed archiving
Current log# 2 seq# 12308 mem# 0: O:\ORACLE\ORADATA\O\REDO02.LOG
The problem seems to be that the oracle service can't start a new logfile when the previous one is full. Does anybody know how I could fix this?
Thanks in advance,
Bram

Hello Maran, thanks for your quick response.
I added the output of archivelog list in the original post, but it's in dutch so here's the (loose) translation of the output:
SQL> archive log list
Databaselog mode Archive mode
Automatic archiving Deactivated
Archive destionation O:\oracle\ora92\RDBMS
Oldest on-line logsequence 12307
Next logsequence for archiving 12307
Current logsequence 12309
Maybe something that can help with the troubleshooting:
The archivelog file is written to O:\oracle\ora92\RDBMS but only AFTER the reboot, I've checked SYSTEM rights to the folder and it has full control.
Babu, thanks aswell for your quick response.
The location exists and space is available but the database doesn't seem to load (that part of) the ora.init file because the output of archive log list still refers to %ORACLE_HOME%\rdbms instead of %ORACLE_HOME%\database\archive.
I've now executed alter system archive log all; and am executing the import where the database usually freezes to see if it still happens. So far I already had this written to the alert_sid.log file after executing the command:
Completed: alter database open
Thu Apr 17 10:00:11 2008
ARCH: Evaluating archive log 1 thread 1 sequence 12307
ARCH: Beginning to archive log 1 thread 1 sequence 12307
Creating archive destination LOG_ARCHIVE_DEST_1: 'O:\ORACLE\ORA92\RDBMS\ARC12307.001'
ARCH: Completed archiving log 1 thread 1 sequence 12307
ARCH: Evaluating archive log 2 thread 1 sequence 12308
ARCH: Beginning to archive log 2 thread 1 sequence 12308
Creating archive destination LOG_ARCHIVE_DEST_1: 'O:\ORACLE\ORA92\RDBMS\ARC12308.001'
ARCH: Completed archiving log 2 thread 1 sequence 12308
And new log files were written to the directory (O:\oracle\ora92\RDBMS) and this time I didn't have to reboot for it to happen. Both were 100mb files without the database actually doing all that much though.
Message was edited by:
user616266

Restore database in archivelog mode

Hello everyone.
I've been looking on google, how to restore my database on archivelog mode, but i did not succeed.
I made my backup using this mode, and now i need to know what is the command need to use to restore this.
Somebody can write the command here please ?
Thanks for help.
Regards,
Dan
ps: Using Oracle 10g XE on Win XP.

user8973191 wrote:
Oh, ok Vijayaraghavan K.
Thx for u help.
And about the users ?
i need to create the "same" user in another machine ?
For example:
In this machine i using the "system" user, where i have my tables.
Another machine, when i restore, my tables will go to the "system" user too? or i can choice? or i need to create one?A true backup is a copy of the data files at the file/block level. The restore is therefore a restore of the data files/blocks. Applying the redo (archive logs) is also done at the block level. None of that knows or cares (or needs to know or care) about logical objects (such as users, tablespaces, tables, rows, etc) within the database. so if you do a proper restore, you are restoring files to a consistent state and thus everything that was defined within those files will be there when restored.

Why we need oracle database reboot to put database in archivelog mode?

I know how to enable archivelog mode, but what is the reason that to enable archive log we need database restart. which process forces to do that.

SantoshGanjure wrote:
Thanks a lot EdStevens for ur prompt response...I need something that satisfies my managers query.
If the question has originated with the PHB, then my flippant answer is even more relevant ...
PHB's are incapable of understanding the technical details, so I prefer to give them the 'executive summary':
"Because that's the way Oracle is written. We can argue over the technical merits of such a requirement 'till the cows come home. But at the end of the day, we have to shut down the database because the 'alter database archivelog' command will only succeed when the database is in mount (not open) mode. Now, let's work out how we can achieve our objective of putting the database in archivelog mode with minimal and appropriately scheduled down time."
As for the actual down time, it's about as minimal as you can get. How long does it take your database to complete a shutdown once the command is given? How long does it take you to type "startup mount"? How long does it take you to type "alter database archivelog"? to type "alter database open;"? For the database to complete the 'open' process? You could even script it so as to eliminate the keyboard typing time.

Put database in archivelog mode

I know how to do it.
using oracle 9.2.0.8 on AIX 5.3
My question is that can I mention a different location
my database installed in /opt/oracle/product/9.2
datafiles/control files and redolof files
in /var/data/oracle/product/9.2/orcl
I want archive to be in
/var/data/oracle/product/9.2
How can I do this
please suggest

Is this standard or enterprise edition?
On the enterprise edition you can use the log_archive_dest_N (n=1..10) instance parameter file to configure up to ten different locations. As an example, you would configure: log_archive_dest_1="location=/var/data/oracle/product/9.2 mandatory" You have to configure log_archive_format as well (unless you want the default format mask); log_archive_format=arch%S.arc and the log_archive_start parameter to true.
Shutdown normal or immediate your database. Once it is shutdown start it up to the mount state:
startup mount
and issue the alter command:
ALTER DATABASE ARCHIVELOG
open your database and you will be working on archivelog mode.
If you are on a standard edition, just proceed as previously indicated, the difference is that log_archive_dest_N parameter is not available, and you can define up to two different file locations, parameters are:
LOG_ARCHIVE_DEST, LOG_ARCHIVE_DUPLEX_DEST
and to indicate mandatory or optional, you'll have to use the LOG_ARCHIVE_MIN_SUCCEED_DEST instance parameter.
~ Madrid

Startup in Archivelog mode

Hi.
I only will do a simply question:
If I have a database in archivelog mode, and ocurr a unexpected shutdown
is There some process that use these logs archived if it´s necesary automatically when I do a normal startup, or only will use the redo logs?
Thanks

Alex Tutor wrote:
Hi.
I only will do a simply question:
If I have a database in archivelog mode, and ocurr a unexpected shutdown
is There some process that use these logs archived if it´s necesary automatically when I do a normal startup, or only will use the redo logs?
ThanksNo ,in this case oracle(SMON) will use only online redo log and will perform instance recovery then database will open.Of course if there will not happen media failure.

Recover datafile in no archivelog mode

Hi, i have always done media recovery using RMAN and this works nicely.
However im doing some testing on a database that does not use RMAN and is not in archivelog mode and im confused.(11.2 - aix)
My testing is as follows;
With the database shutdown cleanly i have taken a copy of dbf file. I then open the database and delete the dbf file.
(to slightly digress how can i still view and update the table from this datafile when the dbf file is dropped?)
In scenario 1 i shutdown the database after dropping the file. I then put the dbf file back and try and startup the database.
I thought that as i have a complete consistent backup i would be able to open the database normally without recovery, but this produced an error saying the file needed media recovery so i recovered the datafile and put online and the the db opened normally.
In scenario 2 i again took a clean backup and removed the dbf file. I then opened my database and switched redo logs 3 times ( i have 3 redo logs). After the 3rd switch the database crashed (which makes sense).
I have tried opening it normally again butagain it says the file needs media recovery. i then tried to recover the file but now i the following error:
ORA-00279: change 6667823 generated at 12/17/2013 11:35:20 needed for thread 1
ORA-00289: suggestion :
/test/oracle/TSTA/TSTA/archivelog/2013_12_17/o1_mf_1_78_%u_.arc
ORA-00280: change 6667823 for thread 1 is in sequence #78
Specify log: {<RET>=suggested | filename | AUTO | CANCEL}
Any help would be greatly appreciated

885842 wrote:
Hi, i have always done media recovery using RMAN and this works nicely.
However im doing some testing on a database that does not use RMAN and is not in archivelog mode and im confused.(11.2 - aix)
My testing is as follows;
With the database shutdown cleanly i have taken a copy of dbf file. I then open the database and delete the dbf file.
(to slightly digress how can i still view and update the table from this datafile when the dbf file is dropped?)
In all *nix systems, the file is still there and open by any processes that had it open when you issued the 'rm' command. It is not really deleted until all processes that have it open have terminated.
In scenario 1 i shutdown the database after dropping the file. I then put the dbf file back and try and startup the database.
I thought that as i have a complete consistent backup i would be able to open the database normally without recovery, but this produced an error saying the file needed media recovery so i recovered the datafile and put online and the the db opened normally.
The backup may have been consistent, but if you restored only that one file, it was several SCN's behind the other files that continued to be updated after you took the consistent backup. "Consistent" means not only consistent interally to the individual file, but consistent with all other data files, plus the control files, plus the redologs.
In scenario 2 i again took a clean backup and removed the dbf file. I then opened my database and switched redo logs 3 times ( i have 3 redo logs). After the 3rd switch the database crashed (which makes sense).
I have tried opening it normally again butagain it says the file needs media recovery. i then tried to recover the file but now i the following error:
ORA-00279: change 6667823 generated at 12/17/2013 11:35:20 needed for thread 1
ORA-00289: suggestion :
/test/oracle/TSTA/TSTA/archivelog/2013_12_17/o1_mf_1_78_%u_.arc
ORA-00280: change 6667823 for thread 1 is in sequence #78
Specify log: {<RET>=suggested | filename | AUTO | CANCEL}
Any help would be greatly appreciated
You said you are not in archivelog mode. So where did this reference to an archivelog come from?

Why archivelog mode is mandatory for hot backups.

Hi,
Why the database should be put in archivelog mode when we take hot backups. What happens actually. Why can't we put in noarchivelog mode and take hot backups?

Suppose you are not using RMAN then you must put your tablespaces in backup mode. Otherwise the backup you can take will be unrecoverable.
In order to recover from a Hot Backup you have to do followings
- Restore your Hot Backup
- Recover the database - here, the archive log files are important
How does Oracle know what has to be recovered from archivelogs?It looks at the SCN.
If you never put your tablespaces in backup mode then SCN would not be frozen and Oracle wouldn't know what has to be recovered.
Talip Hakan Ozturk
http://taliphakanozturken.wordpress.com/

Would you use scripts to take database out of archivelog mode?

Background: we do not use RMAN for this production database and we do cold backups each night.
I've been asked to take one of our databases out of archivelog mode for a few days and then to put it back into archivelog mode.
Being somewhat neurotic about these things - I've only ever done it manually - so that I can make sure nothing goes wrong and to check that there was a last successful cold backup prior to the switch.
However, management have asked that it be done automatically by scripts after the backup.
Whilst I can create a script that will do this - the question I'd like to ask the forum is - have you and would you?

I do not agree with taking any Oracle database out of archivelog mode and I do not approve of management's decision and have told them why and all the risks they run.
This is not a test database - it IS a production database and other users do make changes daily.
What I did not reveal until now is that I have also been asked to take this database out of archivelog mode for 5 days and suspend all cold back-ups for the whole server (with 3 other instances on the server - also impacted) - and have warned them accordingly.
However, management have been swayed by someone who feels they know better - who claims to be a qualified Oracle DBA - I'm OCP but seems to have little sway since this person has convinced management that my 6 years experience and qualification is less than his application experience in GIS.
They want to run an upload which will take around 5 days to do - I'm somewhat miffed that an upload can take that length of time and question whether a GIS system should need such an upload for such a length of time.
Let's no go into the problem with tablespaces and space within the tablespaces - the upload will fail because they do not understand that if the tablespace fills up - the upload will fail - but despite asking - I've been told that there is no problem - there will be a problem - since I refuse to allow autoextend on this database - because of the potential to run out of disk space.
I've explained until I'm blue in the face - never mind that even if the upload is successful - the cold backup tape won't be able to backup the data because it'll be potentially over the limit.
I've even warned of my past experience x 3 - when a RAID controller failed - literally died in front of my eyes - I just happened to be standing next to one of the servers talking to someone - and the whole server was lost - but because we had the archivelogs - and the previous day's backup - we got it up and running after Dell replaced the RAID controller - within 1 hour.
That's the joy of Oracle - nothing is lost other than the last archivelog - depending on the time you set the switch.
I love this system - it's so recoverable - it such a joy to work with.
But then when I get told to suspend cold backups and then at the same time suspend archivelog mode - I go mad.
Don't get me started - I've gone down the usual path of why hire an OCP when you won't listen to what I say, etc.
Management are demanding I switch it automatically out of archivelog mode etc.
I can do it - but my question is still - have you guys?
And why did you do it?
All I can say is that I hope that the server has a RAID contoller failure and finally they listen to me - the chances of that are minimal - so management are saying - do it or else.
Oh and by the way - the GIS team have asked for this to be bi-monthly - automated.
And have said that:
"You can not rollback or recover the loading process using the log files because it involves a far more complex operation that would most likely not be successful."
Nice phrase that to an OCP.
I welcome your comments - as I really am in a hole here.
Thanks.

How ArchiveLog mode impacts performance?

Similar Messages

Maybe you are looking for