Netbackup 8.0 Blueprint Oracle
Netbackup 8.0 Blueprint Oracle
Netbackup 8.0 Blueprint Oracle
2
3
4
5
6
7
8
9
10
11
12
• Every database has one or more physical data files. The data files contain a
physical copy of the user data and the data dictionary, which is the description of
the structure of the user data.
• Within the database, user data is organized logically into tablespaces. A tablespace
is a collection of one or more data files. A data file can only contain data for a
single tablespace.
• A data file is structured into data blocks, which are the smallest units of space
within an Oracle database.
‐ All read and write actions performed on a database are at the block level.
‐ extent is a contiguous set of data blocks obtained in a single allocation
‐ segment is a collection of all extents for a specific structure. There are different
types of
segments, including data segments, index segments, and rollback segments.
13
• Every database has two or more redo log files. The complete set of redo log files
for a database is collectively known as the database’s redo log. The primary
function of the redo log is to record all changes made to the database. If a failure
prevents modified data from being written permanently to a data file, changes are
still available from the redo log. Therefore, there is less likelihood of data loss.
Redo logs are used during recovery operations. The process of applying the redo
log during a recovery operation is called rolling forward.
• Oracle automatically generates Online Redo Logs. You can't have an Oracle
instance without online redo logs.
• Redo log files are archived only when the database is configured in ARCHIVELOG
mode and contain redo entries and are copied to one or more archive destinations
before reuse.
14
• Oracle Basics‐ Additional Files for data protection
‐ Control files
‐ Parameter files
‐ Password file
• The parameter file is often referred to as the init.ora file. Beginning with Oracle9i,
there are two types of parameter files: • A PFILE (Parameter FILE), the traditional
type of parameter file, is an ordinary text file that can be directly modified with a
text editor. • An SPFILE (Server Parameter FILE) is a binary file initially built from a
PFILE using the CREATE SPFILE statement.
• A password file is optional for a database except in cases where remote users
connect through Oracle Net with administrative privileges.
15
16
17
18
19
20
21
22
The Oracle architecture components shown in the graphic on this slide
RMAN can back up data to disk or to a stand‐alone tape drive under its own control,
but it requires the assistance of a third‐party media manager to use a robotic tape
library for backups. NetBackup serves as this media manager. The primary function of
a media manager is to track all of the tapes used by RMAN for backups and to
manage the mounting and un‐mounting of those tapes.
The Oracle server process allocated for a channel invokes the API to open a data
stream. The call is then serviced by NBUO, which locates and mounts a tape.
23
24
25
26
Oracle databases can be backed up with file system based backup policies, however
Oracle policies, with the NetBackup for Oracle agent, are preferred. This is mainly due
to their ability to perform online backups integrated with Oracle’s Recovery Manager
(RMAN).
In NetBackup 7.7 and prior, the NetBackup for Oracle agent supports both stream
based backups and proxy copy backups.
Each method has different behavior and schedules, and the RMAN syntax is different.
27
With stream based backups, RMAN is told to generate a backup data stream from the
corresponding database, which is sent to NetBackup.
A specific NetBackup policy and Application Backup schedule can be targeted by
Oracle by specifying environment variables read by RMAN. The Application Schedule
is used by NetBackup to give RMAN the ability to control the backup timing. If no
specific schedule is specified, than NetBackup looks for any Application Backup
schedule with an open start window.
An Automatic backup schedule can optionally be used to allow NetBackup to
schedule and run the scripts that tell RMAN to initiate these backups, instead of
having RMAN schedule the backups itself.
With proxy copy backups , RMAN generates a list of database files, which are then
passed to NetBackup for backup. Unlike stream‐based backups, this backup type
requires some type of snapshot mechanism, and does not use Application Schedules.
28
Instead, an Automatic schedule is used, which is given by an environment variable. If
the variable is not set, then whatever Automatic Full schedule that is found in the
NetBackup Oracle policy, is used.
Note that proxy copy backups can not be used to backup certain types of database
objects, such as control files, SP files, archive logs, and level 1 differential or
cumulative incremental backups.
Note that Oracle 10g and later do allow proxy copy for Archived Log backups, and
proxy copy can perform incremental backups, but only as Block Level Incremental
backups on UNIX (using Veritas Storage Foundation for Oracle).
28
At the heart of Net Backup’s ability to backup Oracle databases is the linking of
Oracle’s RMAN backup utility to NetBackup. This allows RMAN to back up an Oracle
data base and send the backup data to NetBackup instead of sending the backup data
to a disk file on the Oracle host.
Where NetBackup stores backup data is determined by the NetBackup Policy and
schedule the backup data is directed to. When RMAN is used to backup an Oracle
database to NetBackup, it can specify what master server, policy, and schedule the
backup should be directed to..
Legacy NetBackup Oracle policies have a feature that allow for a NetBackup Master
server to run pre‐existing scripts that invoke RMAN with appropriate syntax to backup
the database and send the backup data to NetBackup. In this Instance, NetBackup is
simply aware of the oracle host and location of the script on the Oracle Host . It is not
aware of the database or databases that the script may backup. When NetBackup
invokes the script, it passes information to the script about the policy and schedule
that caused the script to be invoked. The author of the script can use this information
to invoke the appropriate RMAN syntax to back up the database. These types of
policies do not have to run scripts. They can be also be used to allow DBA’s to run
RMAN interactively to backup a database and send the backup to NetBackup.
Changes in the environment may require new scripts to be created and or existing
29
scripts
29
In NetBackup 7.6, an Oracle policy has three types of client lists that can be chosen
from.
One is a list of hosts, or servers, and determines that this policy is treated the same
as an Oracle policy prior to NetBackup 7.6. These legacy Oracle policies have not
changed, and the available schedule types are still Automatic Full, Automatic
Differential Incremental, Automatic Cumulative Incremental and Application Backup.
Likewise, the types of backup selections for host‐based policies are still either Scripts
or Templates. The selection for these in the console is shown as Client for use with
scripts or templates.
The other two client lists are Instances, or Instance Groups, and both determine that
the policy is considered a new Oracle “Intelligent” Policy, introduced in NetBackup
7.6. These Oracle policies have different schedule types and backup selection lists
that will be discussed in the following slides.
Note that whichever type of client list is desired, a policy can only have one client list
type selected.
30
31
APPLICATION schedule is the equivalent of a USER schedule in File system based
policies and it allows an RMAN DBA to manually run a backup of an oracle database
and send the backup to NetBackup. APPLICATION schedule can only run when one of
it's start windows are open.
32
AUTOMATIC schedules allow NetBackup to control when the backup kicks off.
AUTOMATIC schedules can be used as targets if RMAN is performing a PROXY COPY
backup.
33
The slide shows the NetBackup activity monitor details for stream based Oracle
backups when using Automatic schedules
If using Proxy copy backups there is no Application Backup schedule.
A Control File backup needs an Application schedule as it cannot
be backed up via a Proxy Copy backup type and uses the
STREAM method.
34
35
36
Scripts are used by NetBackup for Oracle to allow the NetBackup scheduler to kickoff
an Oracle backup. When the NetBackup scheduler runs an Automatic schedule of an
Oracle policy and sees a script in the policy’s backup selection list, it executes the
script passing the script information through environment variables, such as those
shown on the slide.
These scripts can be written by hand, but are often generated from templates, which
are discussed on the following slide.
Although templates create scripts that handle simple scenarios, backing up multiple
databases with one script, and other more complex scenarios, can be performed.
>>>Note that this usually requires manual script modifications, and that scripts can
be prone to coding errors, so testing should be performed.
More information on scripts for Oracle backups can be found in the
NetBackup 8.0 for Oracle Administrator's
Guide
http://www.veritas.com/docs/000116380
37
Templates are used to generate command or shell scripts that will invoke RMAN with
correct syntax for the client on which the script is generated, for either stream based
or proxy copy based backups. Backup templates are stored on the master server but
restore templates are stored on the client.
Templates are generated using the Backup, Archive and Restore (BAR) Console, by
running a wizard that leads the user through their creation. Their purpose is to
correctly generate the script run by an automatic schedule to perform the backup.
Bpdbsbora can be used on the client to either generate and run the script, or simply
generate the script to be examined.
When generating a script based on a template, you have the ability to tie the
template to a specific policy, schedule, master server and client, and to a single
database. If such templates are used in other policies, the behavior may be
unpredictable, unless the script is manually modified.
38
39
40
41
Before an instance can be included in a policy for backup, NetBackup needs to know
information about the instance. This information is stored in a central repository that
is part of the NBDB database, not as part of the policy.
There is an auto‐discovery process that will automatically discover instances of Oracle
databases. NetBackup administrators also have the ability to add instances to this
repository either using the NetBackup Administration console or a command line
utility.
The database not only needs to be listed in this repository but it also has to be
registered. Registration is the process of assigning credentials to the instance so that
NetBackup is able to log into the database and back it up. This credential information
can include the operating system login credentials of the host where the instance
resides, the Oracle login credentials, and if a Recovery Catalog is going to be used, the
identity of the recovery catalog and it’s login credentials.
An instance group is a collection of
instances that share a common set of
42
credentials (os, db, Recovery Catalog).
Instance Groups can be manually created to simplify registration.
Auto‐discovered instances can be
automatically assigned to an instance
group.
42
By highlighting the Instance sub tab and clicking the star burst icon, you can add an
instance to the repository. Notice the three pieces of information that is being
collected: The name of the instance, the host it exist on, and the path to the Oracle
software needed by the instance.
Clicking OK will add the unregistered instance into the repository.
Clicking the “Provide Credentials” button allows the administrator to register the
instance at the time the instance is added to the repository. Alternately, the instance
can be registered at a later stage when adding it to an Instance group.
43
Instance groups are used to ease the instance registration process. If you have
hundreds of instances to register but they use all of the same credentials, it could be
a lot of work to specify, for each one of those instances, the same set of credentials
over and over and over again.
To make this registration process more convenient, an instance group can be created.
Each instance group can have a set of Unix Host credentials, Windows Host
credentials, Instance Login credentials and Recovery Catalog credentials. Once these
are defined, all that is needed to register an instance is to associate the desired
instance with the instance group. UNIX‐based instances will use the group’s Unix Host
Login credentials, while Windows‐based instances will use the group’s Windows Host
credentials. The Instance Credentials and Recovery Catalog Credentials will be used
by all instances within the group.
Registration can still be a tedious operation, since instances would still need to be
manually assigned to instance groups. However, you can configure an instance group
as an auto‐registration instance group. Any time auto‐discovery finds and adds a new
instance to the repository, that instance will automatically be registered to the auto‐
registration instance group.
Instance groups are managed via the NetBackup Admin Console
44
or via options to the nboraadm command introduced in
NetBackup 7.6 that allow instance groups to be created, have credentials
associated with them, and assign instances to them.
Some examples of managing instance groups using nboraadm are shown here:
nboraadm ‐add_instance_group
nboraadm ‐list_instance_group
nboraadm ‐delete_instance_group
nboraadm ‐modify_instance_group
nboraadm ‐auto_registration
nboraadm ‐disable_auto_registration
nboraadm ‐register_instance ‐instance_group
For more information on using available options to use with
nboraadm please refer NetBackup 7.7 Command Reference
Guide.
44
When creating an Instance Group you can have it contain up to four sets of
credentials: Windows Credentials, Unix/Linux Credentials, Oracle Credentials, and
RMAN Credentials.
NetBackup will either use Oracle or OS credentials to access the database. If OS
credentials are used, NetBackup will use the OS credentials that apply to the host the
instance is running on. If the Instance group does not contain credentials for the
required OS, then access to the database, and validation, will fail.
Here we see the validation button. In this screen shot, it is greyed out because we are
creating an instance group. Instance group validation occurs against the instances
added to the group, and can only run against instance groups that have instances
associated with them. This button will become active when at least one instance uses
this group for registration.
45
Whether you register the instance initially when you add it to NetBackup OR after it
has been added, the same dialog box is used for instance registration.
One way to register the instance is by associating the instance with an existing
instance group. Instance Groups will all use the same set of
Credentials, so instances that use the same login credentials can
be added to the same Instance group. If the radio button is selected at
the top of the dialog box, the drop down box will be active and you will be able to
select from the existing instance groups.
Alternatively, you can register the instance by using instance credentials, as was done
in this example on the slide.
Instance Credentials will require an operating system username and password, and if
the operating system is Windows, the domain will need to be specified.
The Authentication frame determines what credentials will be used: either the OS
credentials or Oracle credentials as determined by the radio buttons.
The Oracle RMAN Recovery Catalog Credentials frame allows you to specify whether
or not an Oracle Recovery Catalog will be used. Check the box in this frame and fill in
the rest of the information like the username, password, and TNS identifier for the
46
recovery catalog.
Finally, clicking OK will validate the credentials and save the registration.
46
Registration credential validation ensures that the credentials will work for that
instance. This validation occurs when an instance is registered, modified, or added to
an instance group, or when an instance group is modified, and will occur for all
instances in the group, even if only one instance is modified or added.
Registration credential validation failure will identify which credentials failed, to make
troubleshooting easier. If a failure occurs, the registration can be aborted, or be
allowed to continue regardless.
47
When credentials are validated against an instance group, the validation actually
occurs against all the instances associated with that group. A report is generated for
each instance that failed the validation specifying what credentials in the group
(OS/Oracle) caused the validation failure.
Validation of the Group occurs by clicking the OK button when adding an instance to a
group or when an instance associated with the group is changed, or when one or
more sets of credentials of the group are changed.
48
To setup auto‐registration in NetBackup for newly identified Oracle instances, select
either NetBackup Management > Applications > Oracle, or NetBackup Management
> Applications > Oracle > Instances in the object tree, then click Actions > Auto
Registration from the menu bar. This brings up the Automatic Registration property
box. Click on the Automatically Register Newly Discovered Instances, and then select
the instance group you want these instances to join.
49
NetBackup will attempt to automatically discover new instances of Oracle databases.
This process of instance discovery is a recurring operation that runs every 5
minutes (300 seconds) by default. This can be changed by using bpsetconfig
to change the value of NBARS_DISCOVERY_TIMER which specifies the discovery
interval period, in seconds.
Two new processes exist that actually perform the discovery operations.
nbdisco, which runs on both the master server and clients, is responsible for
collecting the information from the clients. It provides a general client discovery
infrastructure.
For example, on Linux clients when looking for Oracle instances, it consults the
/etc/oratab file, so any instances that are not referenced in this file will not be auto‐
discovered and will have to be added manually.
The information that is collected about each instance includes the path to the oracle
installation for that instance, the name of the instance (SID), and the name of the
host where the instance resides.
nbars, which runs only on the master server, is responsible for taking this data,
checking if the instance is already known to the master server, and adding the
information about the instance into the repository in the NetBackup relational
50
database (NBDB).
50
NetBackup maintains an Oracle
instance state. Instances can be
Active or Inactive. Only Active
instances will be backed up. Any
registered instance, regardless of its
state, can be added to a policy.
The state of an instance can be
managed using the NetBackup
Administration Console, or the
51
nboraadm utility introduced in
NetBackup 7.6.
By default, instances are active. An
example would be to make an
instance inactive if you know it will be
made offline for a long period of time,
for example for maintenance, and
don’t want failed jobs while it is
offline.
51
Since Oracle Intelligent
Policies generate the scripts, Oracle
Intelligent Policies have to collect
more information up front than
regular or classic Oracle policies.
This additional data gets stored as
part of the policy definition as
opposed to the script definition as
occurs in classic Oracle polices.
52
By default in 7.6 and later,
NetBackup creates an Oracle
Intelligent Policy, producing an extra
“Oracle” tab and having an
"Instances" tab in place of the
Clients tab.
In NetBackup 7.7.2, OIP produces
“Oracle” and “Instances and
Databases” tab.
To change the policy to a regular or
classic Oracle policy, edit the
"Instances" information to backup
"Clients for use with scripts or
template".
52
53
Host-based Oracle policies have
backup types that include those that
we are familiar with: Full Backup,
Differential and Cumulative
Incremental Backups, and
Application Backup.
>>>Instance-based Oracle policies
introduce the Archived Redo Log
Backup type, and no longer support
54
Application Backups.
>>>The frequencies available for
most schedules are similar to other
policy types, which allow frequencies
as short as per hour.
>>>The exception to this is the
Archived Redo Log Backup, which
allows a frequency granularity down
to minutes.
54
To add an instance or instance group from within the INSTANCES and Databases tab
in the policy, simply click the New button. This will bring up a list of the Oracle
Instances or Instance Groups that are known to the master server.
55
If the client list type is either “instance” or “instance group”, the policy is considered
an Oracle intelligent policy, also known as an instance‐based Oracle policy.
The available backup selections for an instance‐based Oracle policy are different than
legacy, host‐based Oracle policies:
The whole database backs up all datafiles of the database and the database's control
file. NetBackup will also automatically include a second, separate control file backup
because the first control file backup does not contain enough information to perform
restores.
Tablespaces allows for specifically named table spaces to be backed up in the
instances. Again, NetBackup will include a separate control file backup that facilitates
restores.
Datafile allows specific data file paths to be backed up. Again, like the first two types
of selection lists, Datafile selection lists will cause NetBackup to include a separate
56
control file backup.
A Flash Recovery Area is a directory or folder used by Oracle to store archive redo
logs, and possibly used as a default location for RMAN backups. The Oracle DBA can
use RMAN to backup the database to disk, if the instance has a flash recovery area,
and have the backup registered in the database control file.
Having NetBackup perform a Flash Recovery Area backup tells RMAN to make a copy
of the backup that already exists in the FRA, and send the copy to NetBackup. This
new backup will be registered in the database control file as a copy of the first
backup, so RMAN will recognize two copies of the backup. NetBackup will not be
aware of the backup that exists in the FRA, only of the copy that it was sent.
Unlike the other backup selection types, backing up the FRA does not include a
separate backup of the control file.
The last 2 options, Database backup shares and Whole Database‐Data file copy
share only apply to NBU Oracle Copilot feature supported with NBU Appliance
version 2.7.1 running on Appliance 5230, 5330 and later platforms.
56
Changing the backup selection to either Partial Database – Tablespaces or Partial
Database – Datafiles, and clicking Browse, allows you to select from available
tablespaces or datafiles.
In this example we clicked the Browse button and are selecting the ORCL2_TEXT
tablespace found in the ORCL2 instance.
57
The Oracle tab contains Oracle‐specific parameters that, with legacy policies, would
have to be specified in the RMAN script. View the documentation, or click on the
Help button to get more information on these parameters.
New options for bpplinfo have been added that allow for the modification of these
values on the command line.
58
59
60
61
62
63
• The Recovery Wizard solicits information from the user about the desired RMAN
restore and recovery operations to create a template which is saved in a user
specified location.
• Passwords for Oracle database access are stored encrypted in the template and
decrypted at run‐time.
• The restore browser is used to display database objects. A hierarchical display is
provided where objects can be selected for recovery.
Limitations of using Recovery wizard for Oracle restore:
■ The database is displayed only in its current state. If objects have been deleted
from the database since the last backup, these objects do not appear among the
objects you can select for restore. To restore the objects that have been deleted, you
need to restore the entire database point in time before the objects were deleted.
■ Data is restored to the original location. The wizard does not provide a way for the
user to specify alternate file names.
■ The wizard does not restore control files.
64
65
66
>>> If either the Archived or the Online redo logs are missing then you may have to
do a RESETLOGS operation.
>>> A database incarnation is created whenever you open the database with
the RESETLOGS option. After complete recovery, you can resume normal operations
without an OPEN RESETLOGS. After a DBPITR or recovery with a backup control file,
however, you must open the database with the RESETLOGS option, thereby creating a
new incarnation of the database. The database requires a new incarnation to avoid
confusion when two different redo streams have the same SCNs, but occurred at
different times. If you apply the wrong redo to your database, then you will corrupt it.
The existence of multiple incarnations of a single database determines how RMAN
treats backups that are not in the current incarnation path. In almost all cases, the
current database incarnation is the correct one to use. Nevertheless, in some cases
resetting the database to a previous incarnation is the best approach. For example,
you may be dissatisfied with the results of a point‐in‐time recovery that you have
already performed and want to return the database to a time before the RESETLOGS.
An understanding of database incarnations is helpful to prepare for such situations.
>>> OPEN RESETLOG operations
When you open the database with the RESETLOGS option, the database performs the
following actions:
67
Archives the current online redo logs (if they are accessible) and then erases the
contents of the online redo logs and resets the log sequence number to 1.
For example, if the current online redo logs are sequence 1000 and 1001 when you
open RESETLOGS,
then the database archives logs 1000 and 1001 and then resets the online redo logs
to
sequence 1 and 2.
Creates the online redo log files if they do not currently exist.
Initializes redo thread records and online redo log records in the control file to the
beginning of the new database incarnation. More specifically, the database sets
the redo thread status to closed, sets the current thread sequence in the redo
thread records to 1, sets the thread checkpoint of each redo thread to the
beginning of log sequence 1, chooses one redo log from each thread and initialize
its sequence to 1, and so on.
Updates all current datafiles and online redo logs and all subsequent archived redo
logs with a new RESETLOGS SCN and time stamp. Because the database will not
apply an archived redo log to a datafile unless the RESETLOGS SCN and time
stamps match, the RESETLOGS requirement prevents you from corrupting datafiles
with archived logs that are not from direct parent incarnations of the current
incarnation.
To perform recovery through RESETLOGS you must have all archived logs generated
after the most recent backup and at least one control file (current, backup, or
created).
67
68
69
70
71
72
73
More information on protecting Oracle Databases in clustered environments is
available in NetBackup 8.0 for Oracle Administrator's Guide
http://www.veritas.com/docs/000100615
74
>>> To configure a proxy copy Oracle backup, you need to edit the RMAN script and
configure NetBackup for Oracle
NOTE :One of the problems with FILESPERSET=1 is that, for large databaseses, it
causes a lot of "Objects" to be created in deduplicated storage and there is a known
issue with Deduplication and Oracle backups where restores fail because of timeouts
caused by an excessive number of "Objects".
75
76
Some limitations for instance‐based Oracle policies (OIP) are listed on the slide, which
include Oracle 9i no longer being supported, as well as the P‐Linux Oracle client
platform.
Archive Log Backup schedules are stream‐based and do not support proxy copy
operations.
Also, unlike legacy Oracle policy behavior, backups cannot be initiated by running
script commands in RMAN. To allow this behavior, create a host‐based Oracle policy
with an Application Backup schedule with an open start window. In other words,
create a traditional host‐based Oracle policy.
77
78
Much of the logging for a job from an instance‐based Oracle policy is the same as it is
for a host‐based (legacy) Oracle policy, however there are some new tasks that have
some additional logging considerations, as shown on the slide.
79
• The main purpose of bphdb is to run an Oracle Intelligent Policy or a
template or a shell script that calls rman, bporaexp, or bporaimp
• orasbt.dll/libobk
– a shared library module that contains the functions that RMAN can
call
– This library is loaded when RMAN is started.
– The name of this binary depends on the operating system.
80
81
82
In the following slides, an example scenario is used to show screenshots with the
results and output from this scenario.
>>>The Oracle policy used has selected three instances to back up, as shown on the
slide.
>>>The Oracle policy uses a backup selection list to backup only specific listed
tablespaces, instead of backing up the whole database. In this case, the two
tablespaces selected are ORCL_TEST, and ORCL2_TEST. The policy will attempt to
backup both of these tablespaces from all three Oracle instances.
>>>However, the actual tablespaces that exist in Oracle are shown on the slide.
Instance ORCL contains only the tablespace ORCL_TEST. Instance ORCL2 contains only
the tablespace ORCL2_TEST. Finally, instance RMANCAT does not contain any
tablespaces with either of these names.
Let’s see what backup jobs result from this scenario.
83
This slide shows the job hierarchy of our backup job.
>>>First, there is a parent job which manages the other backup jobs, and will
continue to remain in the Running state until all backup jobs have completed.
>>>Next, for each instance, there is a discovery job that figures out what will be
backed up from each instance. Notice that since the RMANCAT instance does not
contains any of the objects from the backup selection list, its discovery job fails with
status code 5400, while the other discovery jobs are successful.
>>>The discovery jobs that succeed spawn additional jobs to perform the actual
backups. Note that by examining the parent job column, we can see the hierarchy of
job creation.
In this case job 114 spawned discovery jobs 115, 116 and 117. Job 115, which is the
discovery job for instance ORCL2, spawns job 118 to backup the found objects (in this
case tablespace ORCL2_TEST) and then spawns job 120 to backup the control file for
instance ORCL2.
Similarly, job 116, the discovery job for ORCL, spawns job 119 and 121 to backup the
found objects and the control file for instance ORCL.
84
This slide shows the discovery job details of job 115. This is the discovery job for the
ORCL2 instance. >>>The details show that the ORCL2_TEST tablespace was found but
that the ORCL_TEST tablespace was not found.
>>>Scrolling further in the job details, you can see the generated RMAN syntax that
will be used to backup the ORCL2_TEST tablespace.
85
This slide shows the details for a failed discovery job. In this case, the Oracle instance
does not contain any of the named objects for this policy, which in this case was
tablespaces with the name ORCL2_TEST and ORCL_TEST. >>>Since the discovery job
could not find any of the objects requested, it returns a 5400 error. Unlike successful
discovery jobs that have backup jobs associated with them, this discovery job has no
backup jobs associated with it.
86
This slide shows backup job 118, which is the job that backs up the tablespace data
from instance ORCL2. The File list shows an Oracle handle created by NetBackup for
this backup job. The format of this handle can be specified in the Oracle tab of the
policy.
>>>The job details, which show information about the backup, are also shown on the
slide.
87
88
89