Missing DBMS_SCHEDULER Jobs via OEM

Under OEM 11.1, have noticed that SYS jobs are not included in the list of current jobs. Of course no DBA would create objects under SYS, however sometimes we do.

If you notice a job is running, even an os command, then query the dba_scheduler_jobs view. This will list all dbms_scheduler jobs, then use notmal dbms_scheduler options to work with this job, prefrebably drop the job.

Invisible Indexes

This is a really nice feature, particulary for me, a development DBA. Essentially you can make an index invisible so the optimizer doesn't use this index. This is excellent testing new indexes. You don't need to spend the time waiting for an index to be recreated each time you want to compare the differences.

To make an index invisable then run:
sql> alter index <index name> INVISIBLE;

To make an index visible:
sql> alter index <index name> VISIBLE;

You can also make the optimizer use all invisible indexes by for the current session by:
sql> alter session set optimizer_use_invisible_indexes = true ;

TAR and Excluding Files

This blog will quickly discuss how to exclude files when using TAR to create an archive.

The syntax to do this is:

tar cvfX <

Packaging Oracle Incidents

Under version 11, oracle introduced an interesting mechanism called "adrci" aka

"AUTOIMATIC DIAGNOSTIC REPOSITORY"

This tool, among other tasks does the following:

 - Automatically purges trace files, alert log files etc
 - Manages "incidents"
 - Packages incidents.

This blog will talk about how to manage the logs and manage incidents. Although many features discussed in this blog can be completed via OEM, I will be discussing how to do this from the command interface.

To Log On, run:

unix> adrci

To show your current environment, run:

adrci> show homes

To set a HOME environment, run:

adrci> set home

To show current incidents, run:

adrci> show incidents

ADR Home = /opt/oracle/admin/diag/rdbms/xxx/XXX:

*************************************************************************
INCIDENT_ID PROBLEM_KEY CREATE_TIME
-------------------- ----------------------------------------------------------- ------------------
12426 ORA 3137 [12333] 2011-11-02 16:47:50.277000 +13:00
1 rows fetched


To purge a particular incident, run:
 
adrci> purge -i n nn nnn
 
To purge a type of data, run:
 
adrci> purge -age nn -type xxx
 
where age is defined in minutes and type can be "alert, incident, trace, dump etc"

11.2 Update Window Details

One of the nice features introduced in recent oracle releases was automated maintenance jobs. This is a nice feature as oracle tries to automate STATS generation and automating execution of advisors.

Sometimes, however the default schedule for these maintenance jobs is not good enough, so to change these windows.

In a nutshell, you need to disable the window, set attributes, duration time, start time etc.

A sample would be:

BEGIN
DBMS_SCHEDULER.DISABLE(
name=>'"SYS"."SATURDAY_WINDOW"',
force=>TRUE);
END;

BEGIN
DBMS_SCHEDULER.SET_ATTRIBUTE(
name=>'"SYS"."SATURDAY_WINDOW"',
attribute=>'DURATION',
value=>numtodsinterval(1080, 'minute'));
END;

BEGIN
DBMS_SCHEDULER.SET_ATTRIBUTE(
name=>'"SYS"."SATURDAY_WINDOW"',
attribute=>'REPEAT_INTERVAL',
value=>'FREQ=WEEKLY;BYDAY=SAT;BYHOUR=3;BYMINUTE=0;BYSECOND=0');
END;

BEGIN
DBMS_SCHEDULER.ENABLE(
name=>'"SYS"."SATURDAY_WINDOW"');
END;

tar / compress in one command

To tar and then compress in one command use the following:

tar cvf - directory | gzip > filename.tar.gz

To uncompress and untar, use the following:

cat filename.tar.gz | gunzip | tar xvf -

rman delete

I have the following scenario. I have database backups I want to delete as I am recreating a database tomorrow morning, long story, but we have found a bug when running a block size of 16k.

I don't want to run the "RMAN DROP DATABASE INCLUDING BACKUPS NOPROMPT" command as the business testers want this environment back reasonably quickly.

What I did was:

unix> rman target /

rman> CONFIGURE RETENTION POLICY TO RECOVERY WINDOW OF 1 DAYS;

rman > DELETE NOPROMPT ALL;

Then it listed all the backup to remove, and I said "YES".

Now when I run the RMAN DROP DATABASE command tomorrow, it will complete in good time, not 2-3 hours it took to delete all my backups.

vi search and replace

Use "%s/searchstring/replacestring/g

If the searchstring or replacestring has a "/", then place a "\" in front. This is an escape character meaning that any character that follows is taken as its literal meaning.

ksh Setting Variables only if the variable is not set

ORATAB=${ORATAB:-/var/opt/oracle/oratab}

DATESTRING=${DATESTRING:-$(date)}

Xauth x11 forwarding on unix

Use the following commands to set pup xauth forwarding on unix using putty.

1. Update putty X11 settings. Click the "ENABLE X11 FORWARDING", for each putty connection.

2. Login to unix server.

3. Run > xauth list

Output will be something like /unix:nn MIT ...

Where nn is a number like 10, 11, 12, ...

4. Start exceed / similar server.

5. Run > export DISPLAY=localhost:nn

If sudoing to another unix account then run the following:

1. Sudo to unix account

2. Run > xauth add unix:nn ... Output from xauth list command ran in step 3 above.

3. Run > export DISPLAY=localhost:nn

Now run your x11 application.

Hostid on solaris

A pretty simple command, just run:

> hostid

Remove unwanted files from FLASH_RECOVERY_AREA

Move backups from the flash recovery area to tertiary storage such as tape.

RMAN> BACKUP RECOVERY AREA;

Note : Flashback logs cannot be backed up outside the recovery area and so are not backed up by BACKUP RECOVERY AREA.

NOTE: Removing these files with an operating system utility will physically remove these files, but the database will still think the files exist and the problem of archival error may still occur. Use the following commands to synchronize the files with the database and use rman to remove these files.

RMAN>CROSSCHECK BACKUP;
RMAN>CROSSCHECK ARCHIVELOG ALL;

RMAN>Delete expired backup;
RMAN>Delete expired archivelog all;
RMAN>Delete force obsolete;

4. Make sure that your guaranteed restore points are necessary. If not, delete them

SQL>Drop restore point restore_point_name

5. If flashback logs are enabled then make sure you have enough space for all the flashback logs. If its not required then you can turn off flashback.

SQL> Alter database FLASHBACK OFF;

6. Review your backup retention policy and if required change the RMAN RETENTION POLICY

RMAN> CONFIGURE RETENTION POLICY TO RECOVERY WINDOW OF 7 DAYS;

Exceptions :

- If RMAN is not part of backup strategy and archivelogs are going to FRA then manual intervention required for deletion of archivelogs. Periodically purse old archivelogs

for example

RMAN> Delete archivelog all completed before 'SYSDATE-7';

- By default RMAN backup goes to FRA. While taking RMAN backup if backup location explictly specified to flash recovery area location then those backup pieces are not considered as part of FRA for auto managment.

- For Archivelogs backup to FRA use USE_DB_RECOVERY_FILE_DEST rather than giving explict path of FRA

SQL> alter system set log_archive_dest_10='LOCATION=USE_DB_RECOVERY_FILE_DEST' scope=both;

- Bug 4911954

Details:
V$RECOVERY_FILE_DEST SPACE_USED and NUMBER_OF_FILES values may be wrong.
eg: Number of files in V$FLASH_RECOVERY_AREA_USAGE and V$RECOVERY_FILE_DEST are different.
This problem can lead to recovery area files being deleted even though there is no space pressure.
Fixed in : 10.2.0.3 , 11.1.0.6


- Bug 5106952

Details:
Flashback logs are not reclaimed by flash recovery area when using guaranteed
restore point (after dropping guaranteed restore point) or when changing
db_flashback_retention_target to a lower value.


11.2 recovery - free up spaceon DB_RECOVERY_FILE_DEST*

Make more disk space available and increase DB_RECOVERY_FILE_DEST_SIZE

SQL> alter system set db_recovery_file_dest_size=xG SCOPE=BOTH; -- (larger amount)

2. Move backups from the flash recovery area to tertiary storage such as tape.

RMAN>BACKUP RECOVERY AREA;

Note : Flashback logs cannot be backed up outside the recovery area and so are not backed up by BACKUP RECOVERY AREA.

3. Run DELETE for any files that have been removed with an operating system utility. If you use host operating system commands to delete files, then the database will not be aware of the resulting free space.

RMAN>CROSSCHECK BACKUP;
RMAN>CROSSCHECK ARCHIVELOG ALL;

RMAN>Delete expired backup;
RMAN>Delete expired archivelog all;
RMAN>Delete force obsolete;

4. Make sure that your guaranteed restore points are necessary. If not, delete them

SQL>Drop restore point ;

5. If flashback logs are enable then make sure you have enough space for all the flashback logs. If its not required then you can turn off flashback.

SQL>Alter database FLASHBACK OFF;

6. Review your backup retention policy and if required change the RMAN RETENTION POLICY

RMAN>CONFIGURE RETENTION POLICY TO RECOVERY WINDOW OF 7 DAYS;

Exceptions :

- If RMAN is not part of backup strategy and archivelogs are going to FRA then manual intervention required for deletion of archivelogs. Periodically purse old archivelogs

for example

RMAN>Delete archivelog all completed before 'SYSDATE-7';

- By default RMAN backup goes to FRA. While taking RMAN backup if backup location explictly specified to flash recovery area location then those backup pieces are not considered as part of FRA for auto managment.

- For Archivelogs backup to FRA use USE_DB_RECOVERY_FILE_DEST rather than giving explict path of FRA

SQL> alter system set log_archive_dest_10='LOCATION=USE_DB_RECOVERY_FILE_DEST' scope=both;

- Bug 4911954

Details:
V$RECOVERY_FILE_DEST SPACE_USED and NUMBER_OF_FILES values may be wrong.
eg: Number of files in V$FLASH_RECOVERY_AREA_USAGE and V$RECOVERY_FILE_DEST are different.
This problem can lead to recovery area files being deleted even though there is no space pressure.
Fixed in : 10.2.0.3 , 11.1.0.6


- Bug 5106952

Details:
Flashback logs are not reclaimed by flash recovery area when using guaranteed
restore point (after dropping guaranteed restore point) or when changing
db_flashback_retention_target to a lower value.