Databases Reference
In-Depth Information
With the first release of Oracle Database 12
c
, applications can take advantage of this
infrastructure using the Oracle JDBC driver as well as the JDBC Universal Connection
Pool. Key components in the architecture include the JDBC replay driver, continuity
director, replay context information, and Transaction Guard.
After an outage occurs, the JDBC replay driver will receive a request from the client
application and send the calls in the request to the Oracle Database. The replay driver
receives directions for each call from the database, and a Fast Application Notification
(FAN) or recoverable error. The replay driver then obtains a new database session and
checks whether the replay can progress. (Where RAC FAN provides notification of
configuration and service level status, connection cleanup will include attaching to a
different live RAC instance where instance failure occurred.) If the continuity directory
requires a replay, then the replay driver resubmits the calls as instructed by the database.
The JDBC driver will relay the response to the client application and it will simply appear
to be delayed if successful. If unsuccessful, the original error is received.
Recovering from Failures and Data Corruption
Despite the prevalence of redundant or protected disk storage, media failures can and
do occur. In cases in which one or more Oracle datafiles are lost due to disk failure, you
must use Oracle database backups to recover the lost data.
There are times when simple human or machine error can also lead to the loss of data,
just as a media failure can. For example, an administrator may accidentally delete a
datafile, or an I/O subsystem may malfunction and corrupt data on the disks. Oracle's
engineered systems containing Exadata Storage Servers help eliminate data corruption
by being fully compliant with Oracle's Hardware Assisted Resilient Data (HARD) ini‐
tiative where integrity of data blocks is checked anytime data is being written to disk.
A useful management feature for non-RAC databases used for diagnosis of on-disk data
failures and repair options is the
Data Recovery Advisor
. This Advisor provides early
detection of corruption through a Health Monitor, failure diagnosis, a failure impact
analysis, repair recommendations, repair feasibility check, repair automation, and val‐
idation of data consistency and recoverability. It can be used with primary, logical
standby, physical standby, and snapshot standby Oracle databases.
The key to being prepared to handle failures of these types are implementing a good
backup-and-recovery strategy and making use of Oracle's growing number of Flashback
capabilities. We'll cover those next.
Developing a Backup-and-Recovery Strategy
Proper development, documentation, and testing of your backup-and-recovery strategy
is one of the most important activities in implementing an Oracle database. You must
