Restore Table Space -- Fields and controls

Backup image page
Containers page
Roll-Forward page
Options page



Backup image page

" " Select an entry from the list
" " Change the media information if necessary
" " Manually enter the backup image information
" " OK
" " Cancel
" " Show Command

" " Select an entry from the list

Select this radio button to use one of the listed backup images to restore the table space or table space partition.

" " Change the media information if necessary

This control, and its associated fields, is only displayed if you select a backup image from the list of available images from the list box above it (that is, Select an entry from the list is selected). If you know that the information displayed in the Media Type column for that backup image is not correct, you can change this information by specifying the correct media of the backup image in the Media Type box. The choices are as follows:

Specify directories or tapes (for NT and UNIX--based systems; Specify directories for OS/2 systems)
If the media type for the backup image is either a directory or tape, correct the path or paths to the backup image in the List of directories box. (This information is read from the recovery history file for the database, or from the recovery history file for the database partition.)

You can specify more than one path to the backup image if the table space or table space partition was backed up to more than one location. In this situation, the backup image will be spread across multiple locations, and all of its portions have the same time stamp. All portions of the backup image must be concurrently available for the restore table space operation to complete.

Tivoli Storage Manager
If the media type for the backup image is Tivoli Storage Manager, use the Sessions box to specify the number of Tivoli Storage Manager sessions that you want. To reset the number of sessions to the default value (1), click on Default.

Vendor DLL
If the media type for the backup image is DLL, specify the DLL name in the DLL Name field, and use the Sessions box to specify the number of DLL sessions that you want. To reset the number of sessions to the default value (1), click on Default.

For single-partition database environments only. To display the list of DLL names, click on Browse.

User Exit (OS/2 systems only)
If the backup image is managed by a user exit, no other information is required.

Note:You cannot use the Control Center to restore backup images that were taken previous to DB2 Version 6.

" " Manually enter the backup image information

This control and its associated fields are only displayed if you do not select a backup image from the list of available images from the list box above it (that is, Select an entry from the list is not selected). Specify the media of the backup image in the Media Type box. The choices are as follows:

Specify directories or tapes (for NT and UNIX--based systems; Specify directories for OS/2 systems)
If the media type for the backup image is either a directory or tape, specify the correct path or paths to the backup image in the List of directories box.

For single-partition database environments only. You can use the Browse push button to list the available directories.

You can specify more than one path to the backup image if the table space was backed up to more than one location. In this situation, the backup image will be spread across multiple locations, and all of its portions have the same time stamp. All portions of the backup image must be concurrently available for the restore table space operation to complete.

Tivoli Storage Manager
If the media type for the backup image is Tivoli Storage Manager, use the Sessions box to specify the number of Tivoli Storage Manager sessions that you want. To reset the number of sessions to the default value (1), click on Default.

Vendor DLL
If the media type for the backup image is DLL, specify the DLL name in the DLL Name field, and use the Sessions box to specify the number of DLL sessions that you want. To reset the number of sessions to the default value (1), click on Default.

For single-partition database environments only. To display the list of DLL names, click on Browse.

User Exit (OS/2 systems only)
If the backup image is managed by a user exit, no other information is required.

The remaining fields for this group of controls are as follows:

Notes:

  1. You do not have to specify the date and time if only one backup image exists at the location specified for the backup image.

  2. You cannot use the Control Center to restore backup images that were taken previous to DB2 Version 6.



Containers page

" " Redirect table space containers
" " Allow new container creation during roll-forward
" " Table spaces
" " Container name
" " Add
" " Change
" " Remove
" " OK
" " Cancel
" " Show Command

" " Redirect table space containers

Click on this check box to perform a redirected restore. When you select this check box, you must select another control on the Containers page to specify the type of container modification that you want to do.

You may not be able to perform a restore operation because a container is in the not valid state. In this situation, you must modify the container. You can either change the container's path or remove the container, then, if you want, perform an add operation to replace the deleted container with another container.

The state of the container is rolled up into that of its table space. To display the container that is in the not valid state, select the table space whose state is not valid. That table space's containers will be displayed in the Container name box, and their states will also be displayed. Select the container that you want, then specify the container modification.
Note: For partitioned database environments only. In a partitioned database environment, if you create a container or change its path, the path for the container must exist at every node.

" " Allow new container creation during roll-forward

Select this check box to allow the information in the Containers field of the Table spaces box to be modified by the roll-forward phase of the restore operation. If any create, remove, or alter container operations are recorded in the database logs, these operations will be reapplied during the roll-forward phase.

Note: For partitioned database environments only. In a partitioned database environment, if you create a container, the path for the container must exist at every node.

If you are not doing roll-forward recovery, this option has no effect.

" " Table spaces

This box displays the table space that is selected for restore. The columns are:

Table space
The name of the table space.

Management
SYSTEM for a system-managed table space (SMS) or DATABASE for a database-managed table space (DMS).

Type
The type of table space. The value can be Regular, Long, or Temporary.

Containers
The number of containers in the table space.

State
The state of the containers in the table space. The value can be:

Valid
All containers have a valid path.

Not valid
At least one container has an invalid path.

If the state of a table space is not valid, at least one of its containers has an invalid path. You cannot restore the table space if a container is in an inconsistent state. In this situation, you must modify the container. You can either change the container's path or remove the container, then, if you want, perform an add operation to replace the deleted container with another container.

To display the container that is in the not valid state, select the table space if its state is not valid. The table space's containers will be displayed in the Container name box, and their states will also be displayed. Select the container that you want, then specify the container modification.

Changed
At least one container was modified during the current session.

" " Container name

This box displays all the containers for the table space that is selected in the Table spaces box. For a system-managed space (SMS), the absolute path and the state of the container are displayed. For a database-managed space (DMS), the absolute path, file size, and the state of the container are displayed. The states are as follows:

Valid
The container has a valid path.

Not valid
The container path is not valid.

You cannot restore the table space if the container is in this state: you must modify the container. You can either change the container's path or remove the container, then, if you want, perform an add operation to replace the deleted container with another container.

Changed
The container was modified during the current session.

Added
The container was added during the current session.

" " Add

Click on this push button to open a container editor window, in which you specify the container to add. The type of container editor window that opens will be the same as that of the table space:

Note: For partitioned database environments only. In a partitioned database environment, if you create a container, the path for the container must exist at every node.

" " Change

Click on this push button to open a window in which you can change the name, drive, and path of the selected container.

For single-partition database environments only. If you select multiple containers, you can only change the drive.

Notes:

  1. If the container state is not valid, you must either change its path, or remove it. If you remove the container, you can add a new container to replace it.

  2. For partitioned database environments only. In a partitioned database environment, if you change a container path, the path for the container must exist at every node.

" " Remove

Click on this push button to delete the selected container.



Roll-Forward page

" " Roll-Forward
" " Overflow directory
" " Leave in roll-forward pending state
" " OK
" " Cancel
" " Show Command

" " Roll-Forward

Click on this check box to perform roll-forward recovery on the table space or table space partition. There are two types of roll-forward recovery:

Roll-Forward to the end of the logs
This is the default. All changes recorded in the logs are applied to the table space.
Note: For partitioned database environments only. If you need to restore a table space partition on a specific database partition you can perform roll-forward recovery on it to the end of the logs to bring it up to the same level as the table space on the other nodes.

Roll-Forward to a point in time
Changes are rolled forward to a specific date and time. If you decide not to specify a date and time, the default is the current date and time. The Next transaction to reapply field displays the date and time of the next transaction that will be reapplied by the roll-forward operation. To determine this date and time, the roll-forward status of the table space is checked. The date and time to which the table space already been rolled forward (if any) is displayed.

Notes:

  1. You can only perform a point-in-time operation on a table space if it contains all parts of all its tables.

  2. For partitioned database environments only. In a partitioned database environment, if you want to roll forward a table space to a point in time, you must restore the table space partition at each database partition that contains a portion of the table space, and roll forward each table space partition individually to the same point in time. This ensures that the table space is consistent at each node.

" " Overflow directory

The roll-forward operation checks for logs in the location specified by the logpath database configuration parameter. If you have moved any logs from the location specified by the logpath parameter, use the Overflow directory field to specify the path to the logs that you moved.

For single-partition database environments only. To display the list of available directories, click on Browse.

" " Leave in roll-forward pending state

Click on this check box to leave the table space or table space partition in the roll-forward pending state at the end of the restore operation. This is the default.
Note:You can only leave the table space in the roll-forward pending state if you roll forward to a point in time. If you specify to leave the table space in the roll-forward pending state and to roll forward to the end of the logs, the table space will not be in the roll-forward pending state when the operation completes. This occurs because there are no additional logs to apply to the table space.

A table space in the roll-forward pending state cannot be used. To turn the roll-forward pending state off and make the table space usable, you can:



Options page

" " Process while
" " Performance
" " OK
" " Cancel
" " Show Command

" " Process while

Select one of the following radio buttons to specify whether the restore operation is to be done offline or online.

Offline
The restore operation occurs offline (that is, no other application or user can be connected to the database or to the database partition when the restore operation is in progress). This is the default.

Online
The restore operation occurs online: other applications and users can be connected to the database or database partition when the restore operation is in progress, but the table space or table space partition itself is not available for use.

" " Performance

Use the following fields to specify the number and size of the buffers to be used for the restore operation. You can increase the number of buffers or the size of each buffer (or both) to cause the restore operation to complete faster. Note that increasing the value specified for either option causes the restore operation to use more memory.

Number of buffers
Type the number of buffers that are used for the restore operation. The default is 2. The valid range is from 1 to 65535.

Size of each buffer in 4KB pages
Type the size of each buffer in 4 KB pages. The value is from the database configuration file. The default is 1024, and the range is from 16 to 524288.