oracle 12c incident,oracle 12c 性特性研究

Oracle cc 12c 官档笔记

1、Thus, it is apparent that Cloud Control 12c can do application-to-disk management,

as it manages everything from the application down to the database, middleware,

virtual machine, server, and storage (disk) level.

2、And in October 2011, during Oracle OpenWorld, the latest Enterprise Manager 12c,

renamed Cloud Control instead of Grid Control, was announced with great fanfare.

3、In 2003, Oracle Database 10g was released—where the " g" stood for  Grid.

4、The National Institute of Standards

(NIST) defines  Cloud Computing as essentially being:

On-demand Access to a Shared Pool of Computing Resources.

在需要访问的计算资源共享池。

5、The five essential characteristics of Cloud computing are defined as on-demand

self-service, pooling of resources, elasticity (able to expand or scale down on notice),

measured service (that is, metering of usage), and broad network access.

自助服务，资源共享，弹性(能够扩大或缩小的通知)，

测量的服务(即，计量使用)，和广泛的网络访问。

6、Setup  | My preferences | Select My Home from

the Cloud Control menu. 设置主页

7、the Cloud Control Summary page ：

You can see which Targets are down

8、It is now easier to manage and diagnose issues with the Enterprise Manager Cloud

Control 12c product itself. The Enterprise Manager console allows you to perform

control operations such as starting, stopping, or restarting the OMS, as well as secure

and resecure operations. You can update the OMS configuration properties, as well

as the Agent properties

9、

10、For example, in the following screenshot, the  Databases page has been

added as a favorite, and also the  Summary page. This was an enhancement request

many customers had made, to allow one to go directly to a certain page that was

used most often.

11、A powerful new feature in Enterprise Manager Cloud Control 12c is Real-Time

Automatic Database Diagnostic Monitor ( ADDM).

A fast non-locking analysis is

performed by this tool to diagnose the database, and this aids troubleshooting

considerably in such situations.

Real-Time ADDM will automatically detect problems like OS-level paging,

database contention, a database hang, and storage subsystem I/O issues, in real

time. Corrective actions will be implemented when possible, otherwise diagnostic

data will be collected to aid in further analysis, when there is no way to fix the

non-working database.

There is another new feature in relation to ADDM, and this is known as Compare

Period ADDM.

12、 Active Session History (ASH)

The wait classes can now be viewed as a tree-map structure, and it is possible to analyze performance data by various dimensions such as top SQL, top sessions, active

sessions, and wait classes. In the earlier versions, you could only have a limited

view of metrics by the dimensions of top Sessions and top SQL

13、For backup purposes, that very important aspect of the DBA's worklife, there is a brand new deployment procedure called system backup. This allows the DBA to

back up multiple databases and files and directories from multiple servers at the

same time. The procedure can back up the database using RMAN and, optionally,

also back up the respective database's home—files and folders are backed up using

Oracle Secure Backup.

14、The  Incident Manager (accessed via Enterprise  | Monitoring | Incident Manager)

is a centralized console to manage all the incidents across the enterprise from

the entire Target subsystem. All incidents can be viewed, managed, diagnosed,

and resolved.

15、This new idea of incidents allows you to concentrate on the high-level incidents

instead of individual events. An incident may be defined as a significant event by

itself, or it can be defined as a combination of events which are all caused by the

same root issue.

For  example, when a Target is down, this can be defined as a significant event.

Whereas, if there is a space issue in the system, there will be related events about

the lack of space from various Targets such as the storage, server, as well as the

database—and all these events can be grouped together as a "lack-of-space" incident.

In Enterprise Manager Cloud Control 12 c, it is possible to create your own incident

rules, and in this way you can combine events together to form a new incident.

16、Automatic Workload Repository  ( AWR), and is stored in the  SYSAUX

Tablespace. Once every hour, by default, the database takes a snapshot of its

workload and statistical information, and stores it in the AWR repository.

This data is saved for 8 days by default.

AWR includes Active Session History ( ASH) as one of its key components.

17、By default, the databases are seen in a Load Map (this can be changed to a Search List, if required). This is a fast, visually color-coded way of finding out the database that is placing the most load on a system (the total active load), or the database that has the most performance issues. Simply placing your mouse cursor on the largest box on the Load Map immediately shows more performance information about

the database, such as the time consumed by database sessions in CPU, I/O, or

WAIT (that is, waits in all other wait class categories), as well as the top three

ADDM findings.

18、Automatic Database Diagnostic Monitor ( ADDM) is like a doctor in the database, a self-diagnostic engine that is built into the Oracle database kernel. ADDM looks for

performance issues and also supplies the recommendations and the fixes. This is part

of the Diagnostics Pack license for Enterprise Manager.

19、For example, if you click on the CPU Wait section in the graph, you will be shown

the exact breakdown of the SQL statements and the sessions in the database that

contributed towards that wait event. You can also move the shaded bar over the time

period and the statements, and sessions displayed in the lower section will change

accordingly. This is displayed in the following screenshot; due to the print size, the

screenshot may not be readable, but it is included for the purpose of illustration:

20、Automatic SQL tuning has been possible from Oracle Database 11g onwards. This

feature can be set up to run automatically in the maintenance window and capture

the highest-impact SQL. This SQL is then analyzed, SQL profiles are generated, and

these profiles can be automatically implemented if they can substantially improve

the execution plans. Thus, the application performance can be improved without

changing the application.

To set up automatic SQL tuning, on the database menu go to Administration |

Oracle Scheduler  | Automated Maintenance Tasks.

The automated maintenance tasks are enabled by default

The default values for these settings are 1200 seconds,

20, and 10,000 respectively

21、Database Replay can be accessed by going to Performance  | Database Replay

on the database menu in Enterprise Manager.

22、Database Replay can capture workloads

on 9.2 or 10.2 database versions, but the actual replay of the workload can only be

done on 11 g versions such as 11.1 (replay was first introduced in 11.1.0.6) and 11.2.

Note that the replay cannot be done on 10g databases (for upgrades of 9 i to 10 g)

because RAT is essentially an 11g product. Hence, there will be no back port for

the replay.

23、This means that SPA can be used to test an upgrade from 9.2 to 10.1 or 10.2, in case you

want to do that, but this is not recommended as you should instead be upgrading to

the latest 11.2 database version (to avoid the extra support costs for older versions,

and to be able to use the new features of the most recent database version)

24、For managing and helping to tune long-running SQL statements, Enterprise Manager

Cloud Control 12c provides a very useful utility called SQL Monitoring, which is

enabled out of the box with no performance impact

25、You can use the  /*+MONITOR*/ hint in your SQL

statements to explicitly ask for SQL monitoring in this case.

You can access the SQL monitoring page by going to Performance  | SQL

Monitoring.

26、This is known as Real-Time ADDM (supported

for any Oracle database Target of version 10.2.0.4 and higher) and is very useful for

situations when the database has become very slow, or has hung and normal ADDM

cannot work since it may not be possible to even log in to the database. Even if it is

possible to log in to the database, taking an AWR snapshot or running an ADDM

task may worsen the problem.

27、

···