This is the second stage in a full-system cold restore of the Informix database if you use ON-Archive as your data recovery tool (that is, for your archives and backups).

• You are logged on as user informix or root .
• The database server is offline.
• If required, you have completed Salvaging Logical Logs for Full-System Cold Restore (ON-Archive). The procedure below assumes you have completed this stage first. You see the following message:

1. Enter the first save set ID for the level-0 archive volume to be used in the restore.

You normally start with the latest available level-0 archive. For more information about identifying the required tapes, see Preparing for Full-System Cold Restore (ON-Archive).

In this example for a parallel archive, you would enter 1 as your first save set ID, corresponding to the ON-Archive request ID that created this save set. For more information about where to find the correct save set ID, see the message Created request #00000001# in Preparing for Full-System Cold Restore (ON-Archive).



The save set ID is usually the same as the request ID of the ON-Archive request that created it. If you cannot identify the ID, enter the following command to determine the IDs of save sets on your tapes:

$ ondatartr 'list/tape=(/dev/rmt/0m)'

The list command might take a long time (especially if the volume contains a continuous backup). You need to perform it for each archive and backup volume. You can interrupt the command with CTRL-C once you see the volume header. To avoid having to use the list command, fully label your tape volumes after archives or backups by writing all relevant information on the tape itself.

The physical restore starts after you enter the save set ID. Each dbspace is restored in sequence starting with rootdbs . The dialogue continues as follows:

Level 0 physical restore started.

The file rootdbs has been retrieved.

The file physdbs has been retrieved.

......... .........

The file psapbtab has been retrieved.

End of level 0 archived reached on this volume.

Does this saveset continue on another volume? (Y/N)

2. Mount all volumes for the save set.

If the save set currently being restored covers more than one volume, enter Y and mount the continuation volume. If there are no more volumes for the save set, enter N and the dialogue continues.

3. Continue with higher level archives of the same dbspace set, if available.

If you have a level-1 archive corresponding to the level-0 archive for the dbspace set already restored, then answer Y to the following prompt. Make sure that you do in fact have the tape available. If you enter Y and there is no tape, you might have to restart the entire restore.

Do you have a level 1 archive to retrieve? (Y/N): Y

Mount the volume with the archived data.

Press the return key when ready.

What saveset ID is to be used on volume/dev/rmt/0m?: 60

Follow the same procedure if you have a level-2 archive corresponding to the level-0 and level-1 archives for the dbspace set already restored.



In a restore from a normal sequential archive (where dbspaceset=* ), the physical restore is finished after you have done steps 1 to 3 once. If you are doing this kind of restore, go straight to step "Bring the database server back up without a logical restore".

4. Restoring other dbspace sets (parallel archive).

1. Check that the following message appears in the message log file:

Recovery Mode

2. Enter a new ondatartr command for each dbspace set. For example, enter the following command:

$ ondatartr 'retrieve/dbspaceset= SET0000072B/tape=(/dev/rmt/0m)'

Dbspace set SET0000072B contains dbspaces not yet restored.

3. Repeat this command for each remaining dbspace set from the archive that you are restoring.

1. Enter the following command:

$ onmode -s

2. Be sure to wait until the system goes into quiescent mode (this might take several minutes) before entering the following command to bring the database server fully online:

Now that you have performed the physical restore the database server goes into fast-recovery mode. You can go on to the next stage, Performing Logical Restore for Full-System Cold Restore (ON-Archive).