Shulou(Shulou.com)05/31 Report--

This article is to share with you what are the common fault categories of oracle. The editor thinks it is very practical, so share it with you as a reference and follow the editor to have a look.

I. several goals of database high availability

MTBF (Mean-Time-Between-Failures)

The average failure time, that is, the frequency of database failures, should be increased as much as possible

Response measures

RAC clustering technology: multiple instances on multiple computers open a physical database to reduce the risk of one or more instance failures

Streams: using advanced queue technology, by parsing the archive log, the archive log is parsed into DDL and DML statements, so as to realize the synchronization between databases. Equivalent to

The replication technology in SQL server can synchronize the database object level or even the whole database without any pressure on the main system.

Data Guard: one primary database, multiple standby databases, the standby database is a transactional consistent copy of the primary database, but when the primary data is down, the

Alternate server to continue to provide services. Equivalent to database mirroring in SQL server.

MTTR (Mean-Time-To-Recover)

The average recovery time and downtime after data failure should be reduced as much as possible.

Minimize data loss

While reducing MTTR, data loss should be reduced as much as possible.

Second, common fault categories

1. Statement-level failure, usually DML,DQL,DCL,DDL, etc.

Invalid data-> data that meets the relevant constraints and satisfies the range of values

Lack of permissions-- > give users the permissions they need

Space allocation problem-- > use alter session enable resumable to increase user quota and table space size

Application logic error-- > fix logic error

two。 User process failure

Abnormal exit of users who are not logged out

Abnormal termination of user session

A program error caused the session to end

For the above errors, the instance background process PMON automatically rolls back uncommitted transactions and releases related lock resources

3. Network malfunction

Listener failure-> configure a backup listener, connect-time failover

Network interface card failure-> using multiple network interface cards

Network connection failure (routing)-- > configure a backup network connection

4. User error

User accidentally deletes or modifies data-> rollback or use flashback query to recover

User deletes table-> flashback deletion from the Recycle Bin

5. Media fault

Disk failure

Disk controller failure

Delete or corrupt data files

Possible solutions, using online transaction logs to restore affected files from backup or restore data files to a new physical location

6. Instance failure

Power load failure

Hardware failure

Background process failed

Abnormal shutdown of database

Solution, restart the instance using startup. Instance realizes automatic recovery, rolling forward committed transactions and rolling back uncommitted transactions according to online log files

Check the alarm log, trace log, etc., to find out the cause of the failure.

3. Restore the background processes related to the instance

1. Checkpoint process (CKPT)

The checkpoint updates the information of the current log file to the data file header, control file, etc.

The purpose of a checkpoint is to synchronize the non-synchronization between the log file and the data file, so that it can be determined that the instance recovery starts from the most recent checkpoint

Trigger conditions for checkpoint

When switching logs

When the database uses immediate,transaction, normal option shutdown database

It is determined according to the setting values of the initial files LOG_CHECKPOINT_INTERVAL, LOG_CHECKPOINT_TIMEOUT and FAST_START_IO_TARGET.

The user is triggered manually

Reasons for checkpoints to exist

Ensure that data that has been modified in memory is written to the data file regularly to prevent data loss caused by system or database failure

Reduce the time required for instance recovery, from the most recent checkpoint to the time of failure

Ensure that all submitted data is written to the data file when the database is closed

Checkpoint information includes

Checkpoint location (in online log files)

System change number (SCN)

Location in the online log file

Log information, etc.

Note: the checkpoint process does not write to the block disk, nor does it write any logs to the online log file

two。 Online log file and log writing process (LGWR)

Online log files record changes in the database

Should be reused to avoid loss.

Log writing process

Triggered on commit

Log buffer / 3 full time

Trigger per second

Finished before DBWn

3. Archiving process (ARCn)

The archiving process is an optional background process

When the database is in archive mode, online log files are automatically archived

All changes made in the database will be fully recorded

The archiving process is initialized each time a log switch occurs, and the online log files are archived automatically before they are reused.

It is recommended to turn on archiving mode.

Fourth, the planning of the three major documents

1. Multiplexing control file

Use at least more than one control file

Separate and release to different disks

Copies of control files should be located on different disk controllers

It is recommended to use RMAN for regular automatic backup

two。 Multiplexing redo log files

At least two members of each log group

Each log member should be on a different disk

Each log member should be on a different disk controller

The members of the group are written at the same time, resulting in a larger Imax O, which should be placed on a high-speed disk.

3. Archive log

The production database should be switched to archive mode

Name according to reasonable archive log

More than one archive path

Thank you for reading! This is the end of this article on "what are the common fault categories of oracle". I hope the above content can be of some help to you, so that you can learn more knowledge. if you think the article is good, you can share it for more people to see!

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.