Article 2 Oracle hardware configuration architecture II: network host structure (organization)

This article is based on 17th articles on Oracle architecture and management.
 
Ii. network host structure
Network host structure, that is, multiple Oracle hosts form an oracle host system. Each host communicates with each other through Oracle Net technology. The network host category structure is subdivided into the following structures:
A. Database Network for remote Query
B. Distributed Database for remote transactions
C. Parallel Database for multiple users to access the same database
D. parallel query operation, used for multiple CPUs as one operation service
E. Customer/Server database applications
F. Web-accessible databases
G. Oracle transport Gateway (Oracle transparent gateway) Access
H. Standby Database
I. Copy the database
J. External File Access

1. Database Network
Oracle uses the Oracle Net technology to allow databases on different hosts to access each other over the network. Each host must run the Oracle Net component. For example, Oracle Net Technology in oracle8 is called SQL * Net or net8 ):
 

2. Remote Query
Database Link is used to identify the data query path for remote o r a C l e database queries. A database link directly or indirectly specifies the host, database, and account used to access a specific object. The database link identifies the accessed host and database by referencing the service name of the database. When a database link is referenced by an S q l statement, o r a C l e opens the Session of the specified database and runs the S q l statement there. Then the data is returned, and the remote session can be enabled for the next use. The database link can be a public link (using d B A to make this link valid to all users in the local database) or a private link.
The following example creates a public database link named h r _ L I n k:
Create public database link hr_link
Connect to HR identified by hrpassword
Using 'uq ';
The create database link command has the following parameters:
A. Optional keyword p u B l I c so that d B A can create links for all users in the database.
B. Link name (h r _ L I n k in this example ).
C. account to be connected (if not specified, use the local user name and password in the remote database ).
D. Service name (h q ).
To use this link, simply add it as a suffix to the table name of the command. The following is an example of using the database link HR _ L I n k to query a remote data table:
Select * from employee @ hr_link where Office = 'annapolis ';
Note that the database link cannot be used to return a value from the L o n g Data Type field.

3. Remote update: Advanced Replication options
Databases using Advanced Replication option can not only query data in a remote database, but also update the database content on the remote host. Updates to a remote database can be combined with updates to the local database into a logical unit of work: Submit or roll back together.
Shows a transaction processing example. One transaction is for the remote host database, and the other transaction is for the local host. In this example, the local e m p l o y e table is updated, the remote e m p l o y e table is located in the database specified by the h r _ L I n k database link, and is updated as part of the same transaction. If an update fails, both transactions must be rolled back. This is achieved by executing the two-phase commit (dual-phase commit) of o r a C l e.

4. Cluster Server
All the configurations discussed so far are for databases accessed by a server. However, according to the hardware configuration, multiple servers may be used to access a database. This configuration is called Oracle Parallel Server (Ops, Parallel Server). In the latest Oracle version, it is called RAC (Real Application Clusters ).

5. Oracle transparent Gateway Access
You can access non-o r a C l e databases from your database. That is, you can create a database link for a non-o r a C l e database service. Then you can query data through the database link, just as the source data is an o r a C l e database.
To access non-o r a C l e data, you must use the Oracle transparent gateway (transparent gateway) product. Each type of accessed database engine requires an independent gateway. The gateway runs on the source host of the accessed data. For example, if the source data is stored in an a s/4 0 0 database, then the Oracle transparent gateway software of a s/4 0 0 is installed on a S/4 0 0 0 server. When executed, the gateway software creates a listener on the source server. If you have a user name and password for the database, you can access specific data objects in the database a s/4 0 0.

6. Backup Database
In o r a c l e 7, another database can be constructed as a "backup" Copy of the primary database. Backup database is a special case for server/server configuration. Each server has a complete copy of the o r a C l e software, and the database file structure is the same (if different, an independent control file should be created for the backup copy ). The two hosts should use the same operating system version and database software version.
In the event of a disaster in the product database, you can open a backup database with almost no data loss-the contents of online redo log files are usually lost. Since the redo log files archived in o r a c l e 8 I are generated through the product instance, they must be copied to the standby system. The standby database remains in the recovery mode until it is used. Once the standby database is opened, it becomes the primary database and cannot be easily reconfigured as the standby database.

Summary: The preceding section briefly introduces various configurable architecture of Oracle. Each of these architectures requires further study. In actual work, you need to consider various situations to select the most appropriate configuration structure or a combination of multiple configuration structures.