Top 30 Oracle 19c feature quick look to make dba life easy

 #1 Dynamic fast_start_failover setting to allow changing standby database without need of stopping and starting the data guard broker. 


 #2 Observe-only mode to discover any failures anticipated during fail over without performing actual transition to available standby database. 


 #3 Propagate restore point to standby which was created in primary database to perform point in time recovery. This allows standby to be restored back to the same state as primary using restore point information kept in control file.


 #4 Standby db flashback when primary flashbacks which helps dba by avoiding overhead task of re-creating standby database.





 #5 IN-Memory Column Store and Data Guard Multi-instance redo apply can be enabled together to provide availability and performance. Fastest redo apply technology of Data guard "Multi instance redo apply" & Fastest analytical query performance by "IN-Memory column store" can be enabled on Active data guard. 

 #6 Clear Flashback Logs based on prediction which are beyond retention period helps when multiple database make use of same flash recovery area. It is useful to keep sufficient storage space available for archive/ undo/ flashback logs etc.

 #7 New parameter to tune Automatic outage resolution within Data guard ; DATA_GUARD_MAX_IO_TIME /  DATA_GUARD_LONGIO_TIME waits for specific time before terminating redo apply process in order to automate outage resolution.




 #8 Finer Granularity Supplemental logging to avoid logging of uninterested tables when supplemental logging is enabled by force at database level.


 #9 Oracle Sharding allows horizontal partitioning of data across multiple non-shared , decentralized databases in order to scale up application & availability based on geographical data distribution. sharding_key will be used within the table for sharding with same structure in all db's and with different set of rows. 

Benefits such as ...

Liner Scalability can be achieved by adding independent databases as a shard db to elastic pool.

Extreme Data availability as unavailability or slowness of particular shard db affects only set of users based on geographical location & not complete application.

Data Sovereignty & Proximity via geographical distribution fulfills the regulatory requirement to keep data within a region and to be available rapidly to users as per affinity.




Shard DB components such as ..

Shard catalog manages automatic deployment to shard db along with generic maintenance tasks & changes without impacting the availability of application. 

Shard Directors acts like listener to connect shard db by routing connection based on shared_key which is built on Oracle Global Service Manager (GSM).

Global Service to access data from any shard database based on role , region affinity , replication log tolerance etc.
  
Connection pools to connect users from client server to shard database based on REST API , OCI or JDBC call.

Oracle 19c autonomous database also supports multiple shard databases as a PDB into a single container database (CDB) but with different shard catalog

#10 Automatic Indexing  to manage indexing such as creating new index , rebuilding existing index to improve sql performance based on supported advisory input.

#11 Big Data & Performance enhancement for IN-Memory which supports external table for ORACLE_HIVE , ORACLE_BIGDATA drivers , Parallel query , Oracle Real Application Cluster , Active Data Guard. And on-demand population of SQL analytics capabilities available for IN-Memory column store to analyze both Internal and External data.

#12 Automatic SQL Plan Management to resolve plan regression by automatically identifying sub-optimal execution plan, evolve based on input from advisory and replace it without human intervention with the help of AWR for prioritization.





#13 Real Time statistics gathering after conventional DML operation to identify optional plan without depending on regular DBMS_STATS job. Also high frequency automatic optimizer statistics can be enabled based on application workload and operations.

#14 Hybrid Partitioning of tables to integrate internal table partitions with external stored files which helps in Big Data SQL of Large size partitions.

#15 DBCA enhancements to createDuplicateDB and relocatePDB to another container database in silent mode ; it also supports createFromRemotePDB by cloning from remote PDB in current CDB.



#16 Oracle Client Image available as extract only software which enables to install oracle client by just downloading the setup file , extracting to oracle_home location and running ./runInstaller

#17 Automatic root script execution by setting up environment and permission to avoid un-interrupted installation end to end.





#18 Dry run validation upgrade for oracle clusterware using gridSetup.sh to pre-validate the checks done during actual clusterware upgrade in order to discover anticipated errors.

#19 Auto upgrade using single command and single configuration file. It performs a pre-check , executes pre-requisites upgrade tasks , actual upgrade , post-checks & post upgrade tasks. Also allows fallback and retry option.

#20 Data Pump support for resource usage using MAX_DATAPUMP_PARALLEL_PER_JOB to control the number of parallel jobs spawned in container ; Default 100 , Range 0-250 , Auto 50% of SESSIONS value

#21 Data Pump single operation for partition allows to combine multiple partitions export or import as a single operation using GROUP_PARTITION_TABLE_DATA parameter.

#22 Data Pump support for wildecards for dump files from object store service based on URL based dump files names which reduces long commands.

#23 Oracle Trace Analyzer support using external SMTP server for notification without depending on local SMTP client to send alerts for monitoring , analysis. It is also integrated with oracle cluster health advisor which on detection of problem triggers trace file analyzer to collect relevant diagnostic file and send email. 

#24 SQL Quarantine feature allows SQL statements which are causing high load or resource consumption on database resources such as CPU/Memory to be terminated and quarantined to avoid getting executed again on db ; relevant execution plan is also quarantines to avoid performance degradation again in future due to same cause. 




#25 Automatic database diagnostic monitoring (ADDM) support for PDB to fine tune database at pluggable database itself.

#26 Secure cluster communication between cluster interconnect which uses single network support to have Transport Layer Security (TLS).It will provide data privacy and integrity between the cluster members. Increase in security is invoked automatically on cluster upgrade to 19c or new installation of RAC without the need of explicit configuration.




#27 Zero Downtime grid infrastructure patching without interrupting database operation through out-of-place patching supported in oracle RAC with minimum 2 nodes using rolling patch. 

#28 Colocation TAG for client routing in which COLOCATION_TAG string value is set in TNS (Transparent Network Substrate) connect_data parameter to route the connection to specific database instance and avoid inter instance communication.  

#29 Dynamic Service Fallback option in which service that is planned for using "preferred" and "available" ; due to instance failure when it moves to "available" instance same will be fall back automatically to "preferred" instance once it rejoins the cluster in order to keep the service running on preferred node most of the time. 

#30 Privilege Analysis available in enterprise edition also which analyzes user privileges and reports its usage/non-usage to remove unwanted privileges. 



No comments:

Post a Comment