Oracle Database 12c: Multitenant, Services and Standard Edition RAC

The installation process of a typical Standard Edition RAC does not differ from the Enterprise Edition. To achieve a successful installation refer to the nice quick guide made by Yury Velikanov and change accordingly the Edition when installing the DB software.

Standard Edition and Feature availability

The first thing that impressed me, is that you’re still able to choose to enable pluggable databases in DBCA even if Multitenant option is not available for the SE.

So I decided to create a container database CDB01 using template files, so all options of EE are normally cabled into the new DB. The Pluggable Database name is PDB01.

As you can see,  the initial banner contains “Real Application Clusters and Automatic Storage Management options“.

Multitenant option is not avilable. How SE reacts to its usage?

First, on the ROOT db, dba_feature_usage_statistics is empty.

This is interesting, because all features are in (remember it’s created from the generic template) , so the feature check is moved from the ROOT to the pluggable databases.

On the local PDB I have:

Having ONE PDB is not triggering the usage of Multitenant (as I was expecting).

How if I try to create a new pluggable database?

A-AH!! Correctly, I can have a maximum of ONE pluggable database in my container.

This allows however:

  • Smooth migration from SE to a Multitenant Architecture
  • Quick upgrade from one release to another

To be sure that I can plug/unplug, I’ve tried it:

Other features of Enterprise off course don’t work

Create a Service on the RAC Standard Edition (just to check if it works)

I’ve just followed the steps to do it on an EE. Keep in mind that I’m using admin managed DB (something will come about policy managed DBs, stay tuned).

As you can see it works pretty well. Comparing to 11g you have to specify the -pdb parameter:

Then I can access my DB (and preferred instance) using the service_name I specified.

 

Let me know what do you think about SE RAC on 12c. It is valuable for you?

I’m also on twitter: @ludovicocaldara

Cheers

Ludo

Oracle Database 12c: RMAN recover at table level

Brett Jordan David MacdonaldOracle Database 12c comes with a new feature named “RMAN table level recovery”.

After a quick try it’s easy to understand that we are talking about Tablespace Point-in-Time Recovery (TSPITR) with some automation to have it near-transparent.

 

How to launch it

The syntax is quite trivial. Suppose you’ve dropped a table ludovico.reco and then purged it (damn!) then you can’t flashback it to before drop and don’t want to flashback the entire database.

 

You can recover the table with:

 

You identify the schema.table:partition to restore, optionally you can pass the pluggable database containing the table to recover, the time definition as usual (scn, seq# or timestamp) and an auxiliary destination.

This Auxiliary destination is well-known to be mandatory for TSPITR. You can pass other options like table renaming or tablespace remapping.

Off course, the database must be open in read-write, in archivelog mode and at least one successful backup must be taken.

How it works

Oracle prepare an auxiliary instance by restoring the SYSTEM, UNDO and SYSAUX tablespaces.

Then it opens in READ-ONLY mode the partial database.

 

It uses then the read-only dictionary to take the tablespace that was containing the table before the data loss. This tablespace (users in my example) is restored and recovered, and the database is opened.

 

At this  point, RMAN starts an export/import with datapump to move the table from the auxiliary database back to the target database:

 

Finally, the auxiliary instance is cleaned:

 

We can check if our table is ok:

 

Oh, and yes, now we can select directly from RMAN! 🙂

 

 My opinion

  • It still needs the amount of space needed to recover the auxiliary instance (system, sysaux, temp and the user tablespace containing the missing data), so it has all the defeats of the typical TSPITR, but it’s automatic so is an improvement for the real life.
  • Restoring the user tablespace separately from the system tablespaces can be an issue if you’re saving backupsets over tape: you can end up by reading twice the same backupset that could be read once instead.

Cheers

Ludovico

It’s time to come back…

My recent move to Switzerland has kept me a little busy.I’ve applied last August for a consulting position at Trivadis, in their Lausanne location. I can’t hide I’m quite excited to be part of a great company, but now I need to “ride the wave” and come back with some new posts, especially now that MySQL and Oracle are out with new releases. (Well, Oracle Database 12c will be out SOON, I hope!).

Stay tuned, some interesting content coming soon.

Ludo

GI PSU 11.2.0.3.1 is installed successfully (with exit code 1)

I’ve just installed the GI PSU 11.2.0.3.1 for a new RAC, I figured out that the command returns 1 even if everithing it’s ok:

The patch is installed correctly, all services are up & running

Script that duplicates a database using a physical standby RAC as source

 It’s possibile to duplicate a database for testing purposes (it’s an example) using a standby database as source. This allows you to off-load the production environment.

This is a simple script that makes use of ASM and classic duplicate, although I guess it’s possible to use the standby DB for a duplicate from active database.
You can launch it everyday to align your test env at a point in time.

Dataguard check script for Real Application Clusters (MAA)

Two years after my posts:
Quick Oracle Dataguard check script and More about Dataguard and how to check it I faced a whole new Dataguard between two Oracle Real Application Clusters, aka Oracle Maximum Availability Architecture (MAA).

This enviromnent is relying on Windows OS. Don’t know how this could be called “availability” but here we are. I revisited my scripts in a quick and very dirty way. Please consider that I did copy and paste to check the alignment once per thread, but it should be improved with some kind of iteration to check each thread in a more structured fashion.

Please foreward me every improvement you implement over my code: it would be nice to post it here.

Clustering the RMAN catalog on a RAC environment

You have your brand new RAC deployed on a cluster and you want to manage your backups through a recovery catalog.
Suppose you don’t have a dedicate server to host your catalog, perhaps you wouldn’t configure your catalog as a RAC database: so why don’t you use Clusterware to configure your catalog as a single instance in cold failover?

OTN has a very nice whitepaper describing how to protect a single instance database. This can be nicely applied on 10g, 10gR2 or 11g: Using Oracle Clusterware to Protect A Single Instance Oracle Database 11g.

Clusterware is appealing also for traditional cold failover clusters. Licensing allows you to use Clusterware as far as you protect Oracle software or 3rd party software that use Oracle as database backend.