Examples

The following sections illustrate how to use archecker commands.

Manually Controlling a Logical Restore

You can manually control the stager and applier using the -l command line option. For Extended Parallel Server, manual control is required when you have multiple storage managers that contain logical logs.

The following examples show how to perform a logical restore. In all examples, the name of the schema command file is cmdfile.

The following example is a typical usage:

archecker -bvs -f cmdfile

This command is equivalent to the following command:

archecker -bvs -f cmdfile -lphys,stage,apply

After the physical restore is complete, the archecker utility starts the stager. After the stager has started, the applier is automatically started.

In the following example, the -lphys option performs a physical-only restore:

archecker -bvs -f cmdfile -lphys

In the following example, the -lstage option starts the archecker stager. The stager extracts the logical log records from the storage manager and saves the applicable records to a table.

archecker -bvs -f cmdfile -lstage

The stager should only be started after physical recovery has completed.

In the following example, the -lapply option starts the archecker applier. It looks in the acu_control table for the transaction to recover. The applier should only be started after the stager has been started.

archecker -bvs -f cmdfile -lapply

Performing a Restore with Multiple Storage Managers (XPS)

If you use multiple storage managers, you can perform a table level restore with archecker by configuring archecker on every node.

To perform a table-level restore that involves multiple storage managers:

1. Create an archecker configuration file on every node.
2. Create a schema command file on every node.
3. Remove old restores by executing the following command on a single node:

   archecker -DX

4. Start the physical restore by executing the following command on each node:

   archecker -bX -lphys

   Do not use the -d option.

5. After the physical restore completes, start the logical restore by executing the following command on each node that contains logical log records:

   archecker -bX -lstage

   Do not use the -d option.

6. After starting all stagers, complete the restore by executing the following command on a single node:

   archecker -bX -lapply

Performing a Parallel Restore

If you have a fragmented table that resides in separate dbspaces, you can perform a physical table-level restore in parallel by executing multiple archecker commands with different schema command files for each dbspace.

In this example, the table is fragmented across three dbspaces. The corresponding schema command files are named cmdfile1, cmdfile2, cmdfile3. The following commands delete previous restores and then perform physical restores on each dbspace in parallel:

archecker -DX
archecker -bvs -f cmdfile1 -lphys
archecker -bvs -f cmdfile2 -lphys
archecker -bvs -f cmdfile3 -lphys

You cannot perform a logical restore in parallel.