ORACLE AUTONOMOUS DATABASE FREQUENTLY ASKED QUESTIONS
GENERAL
What is the Oracle
Autonomous Database?
Oracle Autonomous Database is the
world’s first autonomous data management in the cloud to deliver automated
patching, upgrades, and tuning including performing all routine database
maintenance tasks while the system is running without human intervention. This
new autonomous database cloud is self-driving, self-securing, and self-repairing,
which helps to eliminate manual database management and human errors.
The Oracle Autonomous Database is fully
elastic: You simply specify the number of OCPUs and the storage capacity in
TB's for the database. At any time, you may scale, increase or decrease either
the OCPUs or the storage capacity.
What is the
difference between a Dedicated Infrastructure vs. Shared Infrastructure
deployment?
Dedicated deployment is a deployment
choice that enables you to provision autonomous databases into your own
dedicated Exadata cloud infrastructure, instead of a shared infrastructure with
other tenants.
With Shared Infrastructure deployment,
the simplest configuration, you share the resources of an Exadata cloud
infrastructure. You can quickly get started with no minimum commitment,
enjoying quick database provisioning and independent scalability of compute and
storage.
Currently both deployment options are
available with Autonomous Transaction Processing and Autonomous Data Warehouse.
How do I connect to
the Oracle Autonomous Database?
ADB is built upon the Oracle Database,
therefore, applications and tools that support Oracle Database also support
ADB. These tools and applications connect to the service using standard
database connectivity such as SQL*Net or JDBC.
What is the maximum
number of CPUs and maximum storage capacity that can be provisioned in Oracle
Autonomous Database?
In the current release up to 128 CPUs
and 128TB can be provisioned from the cloud console. Customers requiring more
resources need to call their Oracle account team.
Are there any
limitations on databases features or function in Oracle Autonomous Database?
Not all features present in Oracle
Database Enterprise Edition are available in ADB; for example, database
features designed for administration are not available. Please
see Appendix B in both the Autonomous Data Warehouse and Autonomous Transaction
Processing documentation for the complete list.
Where is Oracle
Autonomous Database available?
The full list of data centers is shown
here: Oracle
Cloud Data Regions for Platform and Infrastructure Services.
What Machine Learning
capabilities come with Autonomous Databases?
Autonomous Databases comes with over 30
machine learning algorithms implemented as SQL functions that leverage the
strengths of the Oracle Autonomous Database. Oracle Machine Learning algorithms perform all processing
inside the Autonomous Database and can mine data tables and views, star schema
data including transactional data, aggregations, unstructured data i.e. CLOB
data type (using Oracle Text to extract tokens) and spatial data. Oracle Machine Learning for SQL functions take full
advantage of database parallelism for model build and model apply and honor all
data and user privileges and security schemes. Data scientists can collaborate
to build, evaluate and deploy machine learning models using Zeppelin based Oracle Machine Learning Notebooks. Oracle Machine Learning models can be included in SQL
queries, BI dashboards and embedded in real-time applications.
What Analytics are
supported in the Autonomous Database?
In addition to the Oracle Machine
Learning algorithms, the Autonomous Database provides an extensive library of analytic and statistical SQL functions. Additionally, insights and
predictions discovered in the Autonomous Database can be further investigated,
analyzed and included in Oracle Analytics Cloud dashboards and other BI tools
and Applications.
What is the price of
Oracle Autonomous Database?
Pricing details can be found on the Autonomous Data
Warehouse or Autonomous
Transaction Processing page of oracle.com.
What are the
provisioning options for Oracle Autonomous Database?
You can provision a License-included
version of ADB, or you can provision a Bring-Your-Own-License (BYOL) instance.
If you wish to use BYOL, then you must apply current database licenses to your
ADB service. The BYOL requirements are described on ADB's pricing page
Note: For BYOL, it is possible to seamlessly transition from 16 or fewer OCPUs
to greater than 16 OCPUs. The only requirement is that the your BYOL must
include RAC when scaling beyond 16 OCPUs. Click here for more
information.
For BYOL, is it
possible to seamlessly transition from 16 or fewer OCPUs to the greater than 16
OCPUs?
Yes. The only requirement is that your
BYOL must include RAC when scaling beyond 16 OCPUs. Click here for
more information.
Where can I get more
information licensing requirements for Oracle Autonomous Database?
This document
describes the licensing requirements for ADB.
What is the minimum
compute requirement for Oracle Autonomous Database?
One OCPU is required to do any work,
but the compute portion of the service instance can be turned off and billing
for compute will be halted. Billing for storage continues as long as the
service instance exists.
What is the minimum
Storage requirement for Oracle Autonomous Database?
One TB.
Do customers need
Exadata licenses to benefit from BYOL?
The only things needed for BYOL are:
Multitenant and when using more than 16 OCPUs, RAC. The standby option (not yet
available) will require Active Data Guard as well.
Can a customer run
Oracle Autonomous Database without the 'Oracle Autonomous Database - Exadata
Storage' line-item?
No, you cannot use any other storage
than Exadata.
However, for Autonomous Database
Dedicated Infrastructure deployments, ‘Oracle Cloud Infrastructure – Database
Exadata Infrastructure’ which includes Exadata storage should be selected
instead of ‘Oracle Autonomous Database - Exadata Storage’.
Can I get a
subscription to Block and Object Storage to be associated to Autonomous
Database and not use Oracle Autonomous Database - Exadata Storage?
No, you cannot use any other storage
than Exadata.
Do you need to
precisely calculate the amount of storage required for a database?
No, you must specify the initial
storage required for your database but ADB is elastic, so it is possible to
grow or shrink your database as needed.
SECURITY
Can customers add new
roles with privileges and assign those roles to a user and/or group of users?
You can create users, roles, etc. as
before. However, note that certain commands are blacklisted, and the list of
these commands is provided in the documentation for each service ADW
and ATP.
Can a customer manage
his or her own encryption keys?
No, in the current version there are no
customer-managed keys. Oracle manages the keys.
Is there a plan for
any 3rd party companies to assess the service for vulnerabilities?
As part of the security approval
process, an external company performs penetration testing.
How does Oracle
Autonomous Database provide data security?
Oracle Autonomous Database protects
against both external attacks and malicious internal users:
All data is encrypted at rest using
transparent data encryption. Network connections from clients to ADB are also
encrypted using the client credentials wallet. Using client credential wallets
includes both server and client-side authentication and provides the highest
level of security.
Oracle automatically applies all security
updates to ensure data is not vulnerable.
Customers are not given OS logons or
SYSDBA privileges to prevent phishing attacking.
Additional in-database features such as
Database Vault, Virtual Private Database and Data Redaction are also available.
Is Data Redaction
available with Oracle Autonomous Database?
Yes, data redaction is part of ADB.
Is it possible to
leverage Database Vault with Oracle Autonomous Database?
Yes, the current version of ADB
includes Database Vault. This can be combined with other PaaS offerings such
Data Safe which is free to use with ADB to create a comprehensive security
layer within ADB. Click
here for more information
If a customer wants
to decommission their Oracle Autonomous Database instance will all data on
disks, archive, backup etc. be safely removed?
The database instance will be dropped
the moment the service is terminated. However, the database is restorable for
up to 60 days, as legally required and documented in the Cloud Hosting and
Delivery Policies. After that the data is gone and cannot be recovered. Click
here for more information.
Is it possible to
setup access control lists (ACLs) in Oracle Autonomous Database?
Yes. it is now possible to specify an
access control list that blocks all IP addresses that are not in the list from
accessing the database. Once an access control list is set that specific
Autonomous Database only accepts connections from addresses in the access
control list and rejects all other client connections. By default, when there
is no network access control list specified the database is accessible from any
IP address. Click
here more information
Is it possible to
control the list of database instances within a wallet?
Yes. On the Database Connection form,
it is now possible to select type of wallet that will be generated:
· Instance Wallet - provides a database-specific
wallet with access to a single database only.
·
Regional Wallet - contains connection details for all ADBs within a given tenant-region.
Note that the instance (database-specific) wallet is the recommended option for
applications where the connection is to only one system. Click
here more information.
(Shared Infrastructure only)
Is it possible to
invalidate a wallet?
Yes. This new wallet rotation feature
makes it easy for customers to invalidate existing client certification keys
for a specific ADW/ATP instance or for all Autonomous Database instances within
a region. Click
here more information.
(Shared Infrastructure only)
Can I store data
covered by HIPAA in Oracle Autonomous Database?
Autonomous Database demonstrates its
HIPAA (Health Insurance Portability and Accountability Act) fitness through an
additional attestation prepared in accordance with AICPA SSAE 18, AT-C sections
205 and 315. For more information visit the following page on cloud.oracle.com
- Oracle Cloud
Compliance and Security.
Can I store data
covered by SOC requirements in Oracle Autonomous Database?
Yes. Autonomous Database has achieved
SOC 1 and SOC 2 compliance. SOC (Security Operation Center) was created because
of the rise of cloud computing and outsourcing of functions to service
organizations. Liability concerns caused a demand in assurance of confidentiality
and privacy of information processed in the cloud.
Does Oracle
Autonomous Database have ISO27001 certification?
Yes. Autonomous Database has achieved
ISO27001 compliance. More information visit the following page on
cloud.oracle.com - Oracle
Cloud Compliance and Security.
Does Oracle
Autonomous Database have ISO27017 certification?
Yes. Autonomous Database has achieved
ISO27017 compliance. More information visit the following page on
cloud.oracle.com - Oracle
Cloud Compliance and Security.
Does Oracle
Autonomous Database have ISO27018 certification?
Yes. Autonomous Database has achieved
ISO27018 compliance. More information visit the following page on
cloud.oracle.com - Oracle
Cloud Compliance and Security.
Does Oracle
Autonomous Database support database links?
Yes, database links are fully supported
in ADB.
For Shared Infrastructure ADB
deployments the target database must be configured to use TCP/IP with SSL
(TCPS) authentication. ATP Dedicated Infrastructure does not support TCP/IP
with SSL (TCPS) encrypted connections for DB Links.
Does Oracle
Autonomous Database support database directory objects?
Yes, directory objects are supported
within ADB. This makes it even easier to migrate existing applications to ADB.
Does Oracle Autonomous
Database support Oracle Graph features?
Autonomous Database has enabled a core
set of functionalities for property graphs.
Does Oracle
Autonomous Database support Oracle Spatial features?
Autonomous Database has enabled the
core set of spatial functionalities including the native spatial type, index,
and associated core analysis operators and functions.
Does Oracle
Autonomous Database support Oracle Text features?
Autonomous Database supports Oracle
Text. This means it’s possible to use standard SQL to index, search, and
analyze text and documents stored in the Oracle Autonomous Database.
Does Oracle
Autonomous Database include Application Express?
Yes. Autonomous Database does include
Oracle Application Express (APEX).
Does Oracle
Autonomous Database include SQL Developer Web?
Yes. Oracle SQL Developer Web is
included with Oracle Autonomous Database.
Does Oracle
Autonomous Database include support for Oracle Rest Data Services?
Yes. Oracle Rest Data Services (ORDS)
is supported by Autonomous Database.
Does Oracle
Autonomous Database support SODA
Yes, Autonomous Database supports
Simple Oracle Document Access (SODA) for REST.
Does Oracle
Autonomous Database support partitioned external tables?
Yes, Autonomous Database supports
partitioned external tables. This provides ability to query multiple data files
in the Object Store as a single external table where the Object Store files can
be represented as multiple logical partitions.
(Shared Infrastructure only)
Does Oracle
Autonomous Database support hybrid partitioned tables?
Yes, Autonomous Database supports
hybrid partitioned tables. This provides ability for a query to access both
internal data and multiple data files in the Object Store as single logical
table.
(Shared Infrastructure, 19c only)
What is the current
availability SLO for Oracle Autonomous Database?
The current SLO for ADB is 99.95%
availability. There is an Extreme Availability option on the roadmap, which
will offer 99.995% availability.
Does the backup
process continue to run even if you shut down the database, for example to
reduce costs?
A full back up will still happen.
However, once the service has been stopped there is no need to do incremental
backups because the database is not active.
What is the process
to restore an Oracle Autonomous Database instance?
The restore process is very simple -
select a specific point in time to restore via the management console.
Can a customer
restore an Oracle Autonomous Database backup into another Database Cloud
deployment?
No, you cannot restore an ADB backup
into another cloud service. You can export an ADB database to an object store
and then import into another service.
Can a customer use an
Oracle Autonomous Database backup to restore a database to another/new Oracle
Autonomous Database instance?
No, ADB backups can only be used to
restore and recover to the same database. However, customers have the ability
to clone a database and can choose to clone either the full database or only
the database metadata.
Oracle Autonomous
Database provides a manual backup option on the cloud console that takes a backup
to the object store. What can I use this back up for? Can I restore it to my
own database server?
No, the manual backup, which you must
move to your object store bucket, can only be used by ADB. You can use the
manual backup functionality for cases like backing up before a large data load.
Note, if you initiate a restore
operation ADB may decide to use your object store backup rather than one of its
automatic backups – if it will be faster to restore.
What happens if you
cannot access Oracle Autonomous Database, i.e. the whole service is down?
Oracle is responsible for the service
availability. Therefore, if the whole service goes down it’s Oracles
responsibility to get the service back online ASAP.
What are the RTO/RPO
(Recovery Point Objective/Recovery Time Objective) SLOs for Oracle Autonomous
Database?
Oracle does not provide RTO/RPO SLOs.
ADB uses the fastest restore method (flashback database, automatic and manual
backups) depending on the point-in-time specified for the recovery operation.
Does Oracle
Autonomous Database provide DR/highly available configuration?
No ADB does not currently provide a
DR/HA configuration. However, it is on the roadmap.
Is it possible to
clone production databases for development/test/QA environments?
Yes, ADB provides the ability to clone
a database and you can choose to clone either the full database or only the
database metadata.
Can the dev/test/QA
cloned PDBs be masked?
Yes, it is possible to use Oracle Data
Safe data masking capabilities with cloned instances of ADB. For more
information click here.
What audit reports
will be available to the customer?
Oracle Data Safe provides a large
selection of audit reports For more information click here.
Does ADB support VPN
and/or FastConnect?
You can connect your on-premises
network to ADB using FastConnect Public Peering and private endpoints
Using Private connectivity to
Autonomous Database Dedicated Infrastructure deployment is supported.
Will clients need to
upgrade their drivers?
Third party drivers must support Oracle
Wallets and certificates. With ADB on Dedicated Infrastructure, both regular
TCP and TCPS (Oracle Wallets and Certificate) connections are supported.
Details on connecting to ATP with via
OCI, ODBC and JDBC can be found in the documentation.
Details on connecting to ADW with via
OCI, ODBC and JDBC can be found in the documentation.
Are private Endpoints
supported?
Private Endpoints are supported with
ADB
Is Oracle Data
Integration via Oracle Cloud Marketplace supported with Oracle Autonomous
Database?
Yes, Oracle data integration tools are
configured to work with ADB.
What are the options
for loading data into Oracle Autonomous Database?
There are multiple ways to load data
into ADB:
ADB is integrated with multiple object
storage services for data loading. You can upload your source files to one of
these object-stores and use the PL/SQL API DBMS_CLOUD to load data into your
database. ADB supports Oracle Cloud Infrastructure Object Storage, Oracle Cloud
Infrastructure Object Storage Classic, AWS S3, and Azure Blob Storage for
loading data. This is the recommended method for loading large data sets. SQL
Developer also provides a data loading wizard that can load data from these
object stores. DBMS_CLOUD to load data is not applicable for ATP-D because
DBMS_CLOUD is not available on ATP-D.
You can use Data Pump Import to load
data from dump files. Data Pump Import is integrated with Oracle Cloud
Infrastructure Object Storage, Oracle Cloud Infrastructure Object Storage
Classic, AWS S3, Microsoft Azure Blob Storage. You can upload your dump files
to one of these object-stores and import from those dump files.
You can also use client-side tools like
SQL*Loader, or your in-house scripts to load data from files residing on your
client machines.
Another option is to use Oracle tools
and cloud services like GoldenGate, ODI, GoldenGate Cloud Service, and Oracle
BI Data Sync.
Any 3rd party ETL tool that can connect
to ADB using the security credentials wallet can also be used to load data.
Do customers need to
upload data files to the Oracle Object Store for data loading?
For large data sets, uploading the
source files to Oracle Object Storage and loading from the Object Store will
most likely be faster than other options.
But you can use the other data loading
options too depending on your use-case and requirements. Click
here for more information.
Is GoldenGate
on-premises and GoldenGate Cloud Service supported with ADB?
Yes, both GoldenGate on-premises and
GoldenGate Cloud Service support ADB as a target system only.
ADB cannot be used as a source system
for GoldenGate. Please see the GoldenGate documentation
for configuring GoldenGate for replication to ADB.
Do I need to gather
optimizer statistics after data loading?
ADB gathers optimizer statistics
automatically if you are using the PL/SQL API DBMS_CLOUD, or if the tools and
scripts you use to load data are using direct path loads. If you are using
conventional DML operations to load data, then you will need to gather
optimizer statistics yourself or let the nightly statistics gathering job
gather statistics for objects that have stale statistics.
Does the data loading
process support Avro files?
Yes. Autonomous Database can now read
Avro files and parse the schema within the file to create the columns with the
appropriate Oracle Database data types. Avro files may include complex types –
like arrays, structs, maps and more; ADW supports avro files that contain Oracle
data types. Click
here for more information.
Does the data loading
process support Parquet files?
Yes. Autonomous Database can now read
Parquet files and parse the schema within the file to create the columns with
the appropriate Oracle Database data types. Click
here for more information.
Does the data loading
process support pre-authenticated URLs?
Yes. If source files reside on the
Oracle Cloud Infrastructure Object Storage, it is possible to use Oracle Cloud
Infrastructure pre-authenticated URIs. When a pre-authenticated request is
created, a unique URL is generated. This unique URL can then be provided to
users, partners, or third parties to access the Object Storage resource target
identified in the pre-authenticated request.
Does ADB come with
any predefined data models?
There are no predefined data models in
the ADB. You create your own schema, tables and other database objects as
needed.
You can use SQL Data Modeler to design
a new logical model and physical model, then generate the required DDL
script(s) to implement your model in ADB. Click
here for more information.
What's the algorithm
for the automatic tuning?
The precise details of how autonomous
capabilities are implemented inside ADB are considered proprietary and
internal.
Can a customer create
primary and foreign key constraints on tables in Oracle Autonomous Database?
Yes, you can create any constraint just
like you do in a regular Oracle Database.
Can a customer create
secondary indexes, partitioned tables, or materialized views?
Yes, you can create secondary indexes,
partitioned tables, or materialized views in ADB.
MANAGING AND MONITORING
Is there an option to
delay or reschedule patches?
For Shared Infrastructure deployments -
No. Oracle patches ADB in maintenance windows. Currently, users cannot change
the patching schedule. The date-time for the next scheduled maintenance is
shown on the OCI console page for each instance.
With Dedicated Infrastructure
deployments, you have the option to change the maintenance windows to control
when the Exadata environments and the databases are updated.
If a quarterly patch
has an adverse effect, who will notice (does Oracle monitor)? Will rollback be
offered?
You cannot rollback patches yourself.
Only Oracle operations can rollback a patch. Oracle operations will monitor
patching and rollback if a patch cannot be applied successfully according to
basic sanity tests. However, there is always a chance that the adverse effect
is only observable from the application. In that case, you will have to notify
Oracle via a Service Request (SR).
Does Oracle
Autonomous Database have auto-scaling?
Yes, ADB on Shared Infrastructure does
have auto-scaling and this enabled by default when you create a new instance.
You can select auto scaling during provisioning or later using the Scale
Up/Down button on the Oracle Cloud Infrastructure console. Click
here for more information
ADB on Dedicated Infrastructure does
not have auto-scaling support.
Can I access the
database server's operating system?
No. ADB does not allow access to the
operating system.
How can I monitor
database performance in Oracle Autonomous Database?
ADB provides the following monitoring
consoles:
Oracle Management Cloud - supports
monitoring Autonomous Databases via its Oracle Database Management console.
This provides both monitoring and alerting for your Autonomous Database
instances. The Performance Hub page provides both high-level and detailed
performance metrics for your Autonomous Database instances. Use of Oracle
Management Cloud is included as part of your Autonomous Data Warehouse and
Autonomous Transaction Processing services. Click
here for more information.
Performance Hub - provides real-time
view of performance data directly on the OCI ADB console.
Performance Hub contains the following:
·
- Time
Range field and time slider
·
- Active
Session History (ASH) Analytics tab
·
- SQL
Monitor tab
These features provide a simple
one-click way to access the same information as found in EM Express, Oracle
Management Cloud (OMC), and SQL Developer Web. The most likely target audience
for this new feature will be cloud DBAs who manage a range of ADB instances
within a single tenant and need a simple and fast way to monitor the
performance of individual ADB instances directly from within the OCI Console. Click
here for more information.
ADB Service Console - is a web-based
service console for each database. Using this console, you can look at performance
metrics like CPU storage utilization and monitor database activity. This
console also provides real-time SQL monitoring for current and past SQL
statements. Click
here for more information.
Can I turn off Oracle
Autonomous Database CPUs to save money?
Yes. You can do this by stopping the
service from the cloud console or from the command line using the OCI APIs.
This operation closes your ADB database, keeps your data in place, and stops
charging for CPUs.
How long does it take
to restart an instance after it has been stopped?
Starting an ADB instance is basically
opening up your database. It is completed in seconds.
Can I create and set
my own resource manager plan for workload management?
No. ADB provides a simple resource
manager plan out of the box.
Can I modify the
resource profiles for existing consumer groups (HIGH, MEDIUM, LOW, etc.)?
Yes. Autonomous Database allows
customers to change the CPU/IO shares for the consumer groups within their ADB
instance (high, medium, low, tp, tpurgent).
This allows for use cases where, for
example, a workload running in the LOW resource group requires a greater share
of CPU/IO resources.
Can I assign specific
database users to specific database services (HIGH, MEDIUM, LOW, etc.) and prevent
them from using other database services?
No. Any user can connect to any
database service of their choice.
Can I use Oracle
Management Cloud to monitor my autonomous database instances?
Yes. Licenses to use the Database
Monitoring features of Oracle Management Cloud are in included in your
subscription to Autonomous Data Warehouse and Autonomous Transaction
Processing.
How can I migrate my
existing Oracle Database to Oracle Autonomous Database?
Since an ADB database has some
restrictions on the object types and Oracle Database Options you need to use a
logical migration method rather than a physical one.
The main migration tool for migrating
to ADB is Data Pump. You can export your schemas and import them into ADB using
Data Pump. To sync up the additional/incremental changes on the source database
during the export/import process you can use GoldenGate or GoldenGate Cloud
Service to replicate those changes to ADB.
In the current release you cannot use
physical migration methods like backup/restore, Data Guard, database clones,
and transportable tablespaces to move your existing database to ADB.
Can I import an existing
Oracle Database schema into Oracle Autonomous Database?
Yes. You can use Data Pump to export
the source schema, move the dump files to the object store, and use Data Pump
Import to load them into your ADB database.
Can I move my
on-premises Oracle Database to Oracle Autonomous Database using RMAN backup and
restore?
No, RMAN restore into ADB is not
supported. You need to use one of the supported migration methods outlined
above.
Can I use original
export import methods (exp/imp) to migrate to Oracle Autonomous Database?
No, the original export/import methods
are not supported in ADB. You need to use one of the supported migration
methods outlined above or in the documentation.
How can I export my
data out from Oracle Autonomous Database?
You can use Data Pump Export to unload
data from ADB. For smaller data sets you can also use any SQL client tool to
spool your data to a text file.
Does Autonomous
Database support XML?
Yes. Oracle XML DB features are now
supported in Autonomous Database. To ensure the security and the performance of
Autonomous Data Warehouse and Autonomous Transaction Processing, some Oracle
XML DB features are restricted so refer to the documentation for a complete
list of supported features.
Does Oracle
Autonomous Database automatically create indexes when needed?
If your ADB instance is using Database
19c then Automatic Indexing is available, however, it is not turned on by
default.
Does Oracle
Autonomous Database automatically create materialized views when needed?
No, in the current release ADB does not
create materialized views automatically.
Can I specify the
amount of memory I want for my database?
No, ADB configures the database memory
(SGA and PGA) based on the number of CPUs you provision. Memory scales linearly
with the number of CPUs.
What is the IO
throughput in Oracle Autonomous Database?
IO throughput depends on the number of
CPUs you provision and scales linearly with the number of CPUs. Scaling the
service is quick and easy in ADB, if you need more IO throughput you can add
more CPUs online in a few seconds.
How does Oracle
Autonomous Database isolate different customers for performance and how does it
make sure a customer's workload does not impact others?
ADB uses Database Resource Manager and
IO Resource Manager to isolate resources for all databases. CPUs, memory, and
IO resources are not over-provisioned in ADB, this makes sure every customer
gets their assigned amount of resources at all times.
Does Oracle
Autonomous Database open a database on multiple nodes or a single node? Does it
depend on the number of CPUs?
ADB may decide to open a database on a
single node or multiple nodes. ADB does not expose these details because it is
not necessary to know about this level of detail.
Where are the
parallel processes running for a parallel query, on a single node or multiple
nodes?
ADB decides where to run a query and
where to spawn the parallel processes, it may be on a single node or multiple
nodes. Customers do not need to know about these implementation details since
ADB controls this automatically.
It is possible to
change the CPU/IO shares for the consumer groups?
Yes, it is possible to change the
CPU/IO shares for the consumer groups within their ADB instance (high, medium,
low, tp, tpurgent). This allows for use cases where, for example, a workload
running in the LOW resource group requires a greater share of CPU/IO resources
than HIGH and MEDIUM.
FAQs for Autonomous
Data Warehouse on Shared Infrastructure
This section includes a list of
frequently asked questions that relate specifically to Oracle Autonomous Data
Warehouse on Shared Infrastructure.
What is Autonomous
Data Warehouse?
Oracle Autonomous Data Warehouse provides
a fully-managed database that is tuned and optimized for data warehouse
workloads.
As a fully managed service, all
database lifecycle operations are managed by the service: the creation of the
data warehouse database, the backups of the database, the patching and the
upgrading of the database, and the growing or the shrinking of the database.
Oracle Autonomous Data Warehouse is
fully elastic: You simply specify the number of OCPUs and the storage capacity
in TB's for the data warehouse. You may scale up or down the number of OCPUs or
the storage capacity at any time with no downtime.
Is it possible to
connect existing tools and application to Autonomous Data Warehouse?
Yes, Oracle Autonomous Data Warehouse
is built upon the Oracle Database, so business intelligence applications and
tools that support Oracle Database also support Oracle Autonomous Data
Warehouse. These tools and applications connect to the service using standard
database connectivity such as SQL*Net or JDBC. Click
here for more information
What is the maximum
number of CPUs and maximum storage capacity that can be provisioned in ADW?
In the current release up to 128 CPUs
and 128TB can be provisioned from the cloud console. Customers requiring more
than these need to call their Oracle account team.
Is Database In-Memory
option available in ADW?
Customers cannot use the features of
the database In-Memory option. Internally, ADW uses some of the key features of
Database In-Memory to optimize performance.
Can a customer use
the machine learning features ADW if they choose the BYOL option but have not
previously licensed Oracle Advanced Analytics?
Yes. Licenses for Oracle Advanced
Analytics are not required for BYOL. However, note that, currently, R
Enterprise is not included or supported with ADW.
Which data models
does ADW support? Is there a recommended data modelling technique for ADW?
ADW supports all the usual schema
models such star schema, snowflake and 3NF. The same basic design concepts
still apply when designing the data warehouse/data mart schema. The main
difference between the ADW and a non-autonomous Oracle Database is that you do
not need to specify the physical properties of tables (e.g., partitioning, indexes,
compression, storage details etc.).
Can a customer create
indexes on tables?
The CREATE INDEX statement is supported
in ADW. However, ADW uses other techniques such as Exadata smart scan and
storage indexes to quickly locate data. ADW will automatically create, maintain
and delete indexes in certain cases such as with enforced and enabled primary
key constraints.
Can users create
primary and foreign key constraints on their tables?
Yes, users can create any constraint
just like they do in a regular Oracle Database.
Does ADW include all
the usual analytic SQL features?
The usual in-database analytic SQL
features are included with ADW.
Can a table be
partitioned? Does the ADW automatically partition tables?
Yes, customers can manually create any
type of partitioned table in ADW. Note that ADB does not provide automatic
partitioning. Instead, ADW creates tables using the column store format which
is automatically managed by the database and provides very fast access to data.
Can the parallelism
settings be changed for queries on ADW?
ADW configures parallelism out of the
box based on the number of CPUs you provision. This means there is no need to
explicitly configure parallelism
But if required you can include
parallelism hints in your queries and enable those hints by setting a database
parameter.
Oracle recommends using the default
settings for optimum performance.
Is it possible to
track user load operations and view the BAD and LOG files?
You have complete control over the
reject limits from within the SQL Developer data loading wizard and using the
command-line approach. This provides control for whether a load fails with the
very first row being rejected or whether a specific number of errors can be
tolerated and possibly reviewed later. The default reject limit is set to zero
to ensure highest level of data completeness. Click
here for more information.
Is support for R and
Python on the roadmap?
Yes. Support for both R and Python is
on the roadmap.
Where can I find more
information on the SQL notebook tool Oracle Machine Learning (OML)?
There are links to tutorials and videos
for OML within the Getting Started section on the OML documentation. Click
here for more information
Oracle ML SQL Notebook contains a range
of example notebooks covering the most commonly used data mining features.
These are all prepared and managed by the Machine Learning product management
team. Additional examples are also available via Github, click here
for more information.
Can an AWS Redshift
database be migrated to ADW?
Yes, SQL Developer provides a migration
wizard for AWS Redshift as it does for other databases. This wizard connects to
the source Redshift database, exports the metadata and the data to AWS S3, and
imports them into ADW. It also allows you to save the migration scripts and run
them outside of SQL Developer anywhere you want. Click
here for more information
Which 3rd party
business intelligence tools are supported on the ADW?
As a general rule, any business
intelligence tool that connects to the Oracle Database using an OCI 'thick'
connection or a JDBC thin connection with support for wallets should work with
ADW. Oracle has partnered with many 3rd party developer to test and certify
tools and connectivity solutions with ADW. Click here for more
information.
FAQs for Autonomous
Transaction Processing on Shared Infrastructure
This section includes a list of
frequently asked questions that relate specifically to Oracle Autonomous
Transaction Processing on Shared Infrastructure.
What is Autonomous
Transaction Processing?
Oracle Autonomous Transaction
Processing is one of a family of cloud services built on the self-driving,
self-securing, and self-repairing Oracle Autonomous Database. Autonomous
Transaction Processing (ATP) enables businesses to safely run a complex mix of
high-performance transactions, reporting, batch, and machine learning along
with simpler and faster application development on the Oracle Database on
Exadata in the cloud.
ATP automates patching, upgrades, and
tuning without human intervention or downtime. Users can instantly create new
ATP databases and easily convert existing databases, dramatically reducing
costs and time to market.
ATP is also fully elastic: you can
instantly and independently, scale the compute or storage, so only the required
resources are provisioned at any given time, decreasing runtime costs.
ATP offers two deployment options:
Shared - With Shared Infrastructure
multiple customers share the resources of a single Exadata Cloud
Infrastructure.
Dedicated - With Dedicated
Infrastructure, a customer's ATP databases are provisioned on a dedicated
Exadata Cloud Infrastructure, with no shared resources, isolated from other
tenants.
What is the maximum
number of CPUs and maximum storage that can be provisioned in ATP?
In the current release up to 128 CPUs
and 128TB can be provisioned from the cloud console. Customers requiring more
than these should contact their Oracle account team.
Are there any
limitations on databases features or function in ATP?
Not all features present in Oracle
Database Enterprise Edition are available in ATP; for example, database
features designed for administration are not available. You can find a complete
list of the features that are not supported in the ATP
documentation.
CONNECTING
How do I connect to
the service?
ATP is built upon the Oracle Database;
applications and tools that support Oracle Database also support ATP. These
tools and applications connect to the service using standard database connectivity
such as SQL*Net or JDBC.
Note, all communications with the
database are encrypted and users are required to download a wallet file after
creating the ATP Shared Infrastructure instance, which contains the necessary
credentials, to establish a connection.
Where should
application servers be provisioned in order to avoid network latency?
Application servers should be
provisioned in the same Oracle Cloud Infrastructure Region.
What pre-defined
database services are available with ATP?
ATP comes with 5 preconfigured database
services, HIGH, MEDIUM, LOW, TP and TPURGENT. The services control the priority
of the sessions when the system is under resource pressure and some control the
parallel degree used. By default, queries will execute serially when connected
to the TP, TPURGENT and LOW services.
Users can specify a parallel degree on
an object or uses optimizer hints to trigger parallel execution to be used when
connected to the TP and TPURGENT services. Queries will be automatically
executed in parallel when sessions are connected to the HIGH and MEDIUM
services. There is also a limit on the number of concurrent sessions that can
execute on the HIGH and MEDIUM services.
Can a customer create
primary and foreign key constraints on tables in ATP?
Yes, customers can create any
constraint just like they do in a regular Oracle Database.
Can a customer create
secondary indexes on tables in ATP?
Yes, customers can create both B-Tree
or Bitmap indexes on any table.
Does ATP
automatically create indexes?
Yes, where the database version is 19c
which is required for Auto-Indexing. Note that auto-indexing is not on by
default.
Can a table be
partitioned in ATP?
Yes, customers can create any type of
partitioned table.
Does ATP
automatically partition tables?
No, not in the initial release of the
services.
Can a customer create
a materialized view in ATP?
Yes, customers can create any
materialized view.
Does ATP
automatically create materialized views?
No, not in the initial release of the
services.
Does ATP have full
PL/SQL support?
Yes, customer can utilize nearly all
the PL/SQL capabilities. The small set of limitations is listed in the
documentation,
FAQs for Autonomous
Database on Dedicated Infrastructure
This section includes a list of
frequently asked questions that relate specifically to Oracle Autonomous
Database on Dedicated Infrastructure.
What is Autonomous
Database on Dedicated Infrastructure?
Autonomous Database on Dedicated
Infrastructure is a deployment choice that enables customers to provision
autonomous databases into their own dedicated Exadata cloud infrastructure,
instead of a shared infrastructure with other tenants. Dedicated deployment
provides the customer customization over operational polices, including
software and hardware isolation for the highest levels of performance and
security governance; it is well suited for customers wanting to deploy Oracle
Database in cloud with common enterprise lifecycle controls and especially
ideal for customers who want a private self-service capability via their own
private database cloud within the public cloud.
What is the
difference between Autonomous Database on Shared Infrastructure and Autonomous
Database on Dedicated Infrastructure?
Autonomous Database cloud service now
offers two deployment choices, shared (ADB-S) and dedicated (ADB-D). With a
shared infrastructure deployment, the simplest configuration, multiple
customers share the resources of an Exadata cloud infrastructure. These
customers can quickly get started with no minimum commitment, enjoying quick
database provisioning and independent scalability of compute and storage.
With a dedicated deployment, the
customer must first subscribe to a dedicated Exadata cloud infrastructure that
is isolated from other tenants, with no shared processor, memory, network or
storage resources. This deployment choice offers greater control of the
software and infrastructure lifecycle, customizable policies for separation of
database workload, software update schedules and versioning, workload
consolidation and availability policies. Both deployment options ensure high
availability, exceptional performance, and multi-level security.
When should I use
ADB-D versus ADB-S?
ADB-S offers an ultra-simple and
elastic database deployment in the cloud, it enables customers to quickly
provision their own autonomous databases and become productive with new
application development. ADB-S delegates all operational decisions to Oracle
for the highest level of Autonomous experience, think of an autonomous vehicle
with no need for a steering wheel or cruise control.
ADB-D, on the other hand is more like
an autonomous vehicle that still includes a steering wheel and cruise control.
ADB-D offers greater control and isolation starting at the Exadata cloud
infrastructure level, with customizable maintenance schedules, software update
versions, backup retention periods, etc. ADB-D allows customers to group and
separate out databases based on organizational structure and criticality of
application workload.
ADB-D is especially ideal for customers
looking to deliver a self-service database capability within a private database
cloud environment running on the public cloud. Using this model, customer fleet
administrators can use the steering wheel and cruise control (customized
operational policies) to tailor an operational plan for different parts of
their organization, then the organization consumes via a self-service
experience similar to the vehicle with no need for a steering wheel or cruise
control.
What types of database
workload are best to run on ADB-D?
You can run all your transaction
processing or data warehouse or mixed workload databases of any size,
scale or criticality on ADB-D. As well as supporting application databases that
may require highest governance, consistent performance and operational
controls.
Are there any
limitations on database features or functions in ADB-D?
Not all features present in Oracle
Database Enterprise Edition are available in ADB-D; for example, database
administration features are not available. Please see Autonomous
Transaction Processing documentation and Autonomous
Data Warehouse documentation for the complete list.
What are the sizes of
the Exadata Cloud Infrastructure racks that are supported on ADB-D?
Both Exadata Infrastructure Quarter and
Half Racks are supported. Full Rack (& Base system) will be available in
future ADB-D releases. See the documentation, Characteristics of Autonomous Exadata Infrastructure Resources,
for details about the capacities and characteristics of these systems.
How many Autonomous
Container Databases and Autonomous Databases are supported on an Exadata Cloud
Infrastructure Quarter Rack?
You can have up to 12 autonomous
container databases with a maximum of 200 autonomous databases on each
container database for the high availability configuration.
What is the role of a
Fleet Administrator in ADB-D?
Fleet administrator is a logical
service role (managed by IAM privileges) responsible for the creation and the
management of Autonomous Exadata Infrastructure (AEI) and Autonomous Container Database
(ACD) resources for dedicated deployments. These resources must be created
before Autonomous databases can be created.
The logical role is materialized by
granting IAM privileges to a group of users (a Fleet Administrator Group) to
manage the AEI and ACD resources. Please see Fleet
Administrator’s Guide to ADB on Dedicated Infrastructure for more
information.
What is the role of a
Database Administrator in ADB-D?
Database Administrator is both a
logical service user role (managed by IAM privileges) and a Database User role
responsible for the creation, monitoring and management of Autonomous
databases.
In order to manage the Autonomous
Database resource, the service role must also be able to at a minimum USE
Autonomous Container Database resources, which must be visible to assign as the
target for the ADB during provisioning.
Database Users are in essence
application developers, users who can connect to and use an Autonomous Database
to store and access the data generated by applications. Database Users are not
managed by IAM like service users, rather they are managed by capabilities
built natively into the Autonomous Database.
A Database Administrator is an IAM
service user who is also a special case Database User, a user with the highest
level of ADB privileges. The Database Administrator can create, connect and
manages schema and user access to autonomous databases. The Database
Administrator has the exact same responsibilities in ADB-D as in ADB-S.
PRICING AND LICENSING
How is ADB-D priced?
There are two components:
1. Subscription to Exadata Cloud
Infrastructure, the minimum term is 48 hours.
2. Database software usage, subscribed
to hourly per active OCPU consumed. Billing happens monthly at an aggregate
level, total of all active OCPUs, for each Exadata Cloud Infrastructure
resource.
ADB dedicated customers can deploy a
mix of both ADW-D and ATP-D instances on the same Exadata Infrastructure rack.
Pricing details can be found on the
Autonomous Database pricing pages on oracle.com/database (ADW/ATP)
Aucun commentaire:
Enregistrer un commentaire