Saturday, March 19, 2022

Oracle Database 21c New Feature: Hybrid Scans with In-Memory

Oracle Database 21c New Feature: Hybrid Scans with In-Memory.

This blog post by Andy Rivenes Product Manager at Oracle highlights this exciting new feature in 21c to the already expansive array of features boasted by Oracle Database In-Memory.

Here is a brief summary.
  • Columns are excluded from being populated to save space. If one or more columns of a table are never accessed by analytic queries then, there is no point in using the space in the IM column store to populate them.
  • In previous versions, the problem with excluding columns was that, you had to make sure that none of the excluded columns were accessed by your analytic queries or else performance for those queries suffered because the object had to be accessed from the row store.
  • Now that In-Memory Hybrid Scans are available in 21c, even if the columns of a table have been excluded and are then part of some analytic queries there usually won't be a huge drop off in performance.
  • The cost of running the query in the IM column store and then retrieving a large number of column values from the row store may not be any better than just accessing the data in the row store.
  • The query will run slower than if all the columns were populated, but in many instances, it will still run much faster than, if the entire table has to be accessed from the row store. 
  • If you are only returning a relatively small number of values then the query can run almost as fast as if all the columns were populated. 
  • There is a tradeoff but the space savings can be quite dramatic.
Enjoy this blog post announcing this exciting new capability now available for Oracle Database In-Memory.

Cheers.

Friday, March 4, 2022

Step-by-Step: How to migrate from AWS RDS to Oracle Autonomous Database via Data Pump

Step-by-Step: How to migrate from AWS RDS to Oracle Autonomous Database via Data Pump.

There is a very good blog post by SATYABRATA MISHRA Senior Product Manager at Oracle on How to migrate from AWS RDS to Oracle Autonomous Database via Data Pump.

Here is a quick recap of the steps involved.
  • Step 1 - Identify the Source Oracle database on AWS RDS.
  • Step 2 - Prepare target Autonomous Database on Oracle DB Cloud.
  • Step 3 - Validate compatibility using CPAT (Cloud Premigration Advisory Tool).
  • Step 4 - Start the export of the source schema.
  • Step 5 - Create an Amazon S3 bucket to upload the data pump dump file. 
  • Step 6 - Import the data pump dump file from the  Amazon S3 bucket to Oracle Autonomous Database.
Enjoy this blog post on How to migrate from AWS RDS to Oracle Autonomous Database via Data Pump.

Cheers.

Saturday, February 12, 2022

Quick Intro: Database Tools in Oracle Cloud Infrastructure (OCI)

Quick Intro: Database Tools in Oracle Cloud Infrastructure (OCI)


There is a very good blog post by Brian Spendolini Product Manager at Oracle on Database Tools Service in Oracle Cloud Infrastructure (OCI).

Here is a quick recap.
  • In November 2021, the Oracle Database Development Tools team launched a service in OCI. This service, called  Database Tools enabled the customerto create a reusable resource for connecting to cloud databases via REST enabled SQL by consolidating all of the pieces of information you need to have to connect to a database.
  • No need to remember which database on which server or remembering passwords for database connections.
  • Added ability for using OCI policies and defining dynamic groups on who can use these connections.
  • Securely store the hostname, port, database name, username, and password creating a reusable, portable resource that can be utilized across clouds, development tools, services, and pipelines.
Enjoy this Quick Intro blog post on the Database Tools service in Oracle Cloud Infrastructure (OCI).

Cheers.

Wednesday, January 19, 2022

Automatic failover for Autonomous Data Guard enabled Autonomous Databases on Exadata Cloud at Customer and Dedicated Infrastructure

Quick Into: Automatic failover for Autonomous Data Guard enabled Autonomous Databases on Exadata Cloud at Customer and Dedicated Infrastructure

There is a very good blog post by Ranganath Srirangapatna Ramachandra Principal Product Manager at Oracle on Automatic failover for Autonomous Data Guard enabled Autonomous Databases on Exadata Cloud at Customer and Dedicated Infrastructure.

Here is a quick recap.
  • Automatic failover AKA Fast-start Failover in Oracle Data Guard.
  • Automatic Failover allows Autonomous Data Guard to automatically fail over to your standby Autonomous database in the event of an outage on your primary database.
  • Automatic failover rapidly fails over the standby Autonomous database to the primary database role without requiring any manual steps.
  • Automatic failover is optional and can be enabled/disabled on your Autonomous Container Databases with Autonomous Data Guard.
  • Both maximum performance and maximum availability protection modes are available with automatic failover. 
  • Automatic failover guarantees zero data loss in Maximum Availability mode.
  • Automatic failover ensures no data is lost that is more than the amount of data (in seconds) specified by the FastStartFailoverLagLimit configuration property in Maximum Performance mode. 
  • FastStartFailoverLagLimit is set to 30 seconds and applies only to maximum performance mode. Automatic failover is only possible when the configured data loss guarantee can be upheld.
  • Oracle automatically troubleshoots and repairs the failure conditions related to the primary database and reinstates it to the standby database role after an automatic failover occurs.
  • You can perform a switchover operation to change the database roles to the original state after the failed primary is reinstated to the standby database.
Enjoy this blog post on Automatic failover for Autonomous Data Guard enabled Autonomous Databases on Exadata Cloud at Customer and Dedicated Infrastructure.

Cheers.

Saturday, January 8, 2022

Creating Non-CDB Database in ExaCS - Step by Step

This is a short step-by-step using dbaascli on how to create an Oracle Non-CDB database in Oracle Exadata Cloud Service (ExaCS). Here, we will see high-level steps with example to create 19c Non-CDB Database in ExaCS 

There is a very good blog post by Sathis Muniyasamy Principal Cloud Architect at Oracle detailing the following steps.
  • List Available Images
  • Download the Image
  • Create DB Home
  • Create Database
  • Verify the Database
  • Viewing the Non-CDB Database and Home in OCI Console
Enjoy the Oracle Non-CDB database in Oracle Exadata Cloud Service (ExaCS).

Cheers.

Friday, December 17, 2021

Oracle Database 21c now supports Enhanced In-Memory External Tables

Oracle Database 21c now supports Enhanced In-Memory External Tables.

This blog post by Andy Rivenes Product Manager at Oracle highlights this exciting new feature in 21c to the already expansive array of features boasted by Oracle Database In-Memory.

Here is a brief summary.
  • With In-Memory 21c, external tables allow an application to populate data in the IM column store that exists outside Oracle Database without the need to to first having to materialize the data within the database.
  • In-Memory 21c allows an application to take advantage of Oracle Database's rich feature set to query and analyze that data and perhaps combine it with other application data that resides within Oracle Database. 
  • With Oracle Database 21c, Database In-Memory provides partition support for that external data to allow even more flexibility in managing fast analytics on many different types of data.
Enjoy this blog post announcing this exciting new capability now available for Oracle Database In-Memory.

Cheers.

Wednesday, November 10, 2021

Oracle Database 21c New Feature: Full Text Columns with In-Memory

 Oracle Database 21c New Feature: Full Text Columns with In-Memory.

This blog post by Andy Rivenes Product Manager at Oracle highlights this exciting new feature in 21c to the already expansive array of features boasted by Oracle Database In-Memory.

Here is a brief summary.
  • In the previoius versions, Oracle Database In-Memory did not support predicates for non-scalar documents e.g. if you were using domain indexes for Oracle full text index, XML Search Index, or JSON Search Index then these were not populated in the IM column store. This resulted in degradataion of performance for searching non-scalar types of data. 
  • Now in Oracle Database 21c In-Memory, there is new support  to store, or populate, these types of domain-specific indexes.
  • The new 21c In-Memory Full Text Columns feature provides a faster way to perform searches on text type data.
  • Full Text Columns do not require maintaining a separate Oracle Text index.
  • Full Text Columns now support various types of application workloads within the Oracle converged database.
Enjoy this blog post announcing this exciting new capability now available for Oracle Database In-Memory.

Cheers.