Restoring Control Data
If the data lost was the POOLDEF file, control minidisk, or any minidisk in the catalog storage group (storage group 1), you will have to do a restore of the file pool control data.
If your data loss affects more than one item, restore the logs first, then the control data, and, finally, the user data.
To restore the file pool control data, you must have the current backup of it. If you do not have a current backup, or if the backup itself is unusable (for example, the tape breaks), follow the instructions in Restoring a File Pool by Generating It Again.
- Log on or reconnect to the server machine. If the server is currently processing in multiple user mode, reconnect to it and enter STOP NOBACKUP. You do not want a control data backup started if you think your control data may be corrupt.
- Add the RESTORE startup parameter to the DMSPARMS file for the
server machine. (For more information about specifying startup parameters,
see File Pool Server Startup Parameters.)
For best performance when restoring the control data, you should also specify a very large catalog buffer pool. To increase the size of the catalog buffer pool, specify the CATBUFFERS startup parameter. For example, you might enter:
CATBUFFERS 5000 - Enter a FILEDEF command to associate the backup file with the ddname
RESTORE. If, for example, your restore file is on tape, you might
enter:
filedef restore tap1 sl volid nnnnnnYou must use SL on the FILEDEF command to indicate IBM standard labels.
If, on the other hand, your backup file is on DASD, you will first have to access the minidisk or the directory. For example, you might enter:access sysadmin:vmsysu.bkup b filedef restore disk backup file bThe file pool server uses a block size of 28672. If you specify a block size (BLKSIZE) in the FILEDEF command, it is ignored. The file identified by ddname RESTORE should be the same file identified by ddname BACKUP during a previous control data back up.
- Enter the FILESERV START command.
A message will ask whether you really want to restore the control data. If all is well, reply 1 for yes.
The file pool server will restore the control data from the backup file associated with ddname RESTORE. When the restore is complete, the server applies all the changes recorded on the repository log minidisks, which brings the file pool to the state it was in at the time of the error. Then the file pool server allows regular multiple user access.
When the restore completes, multiple user mode access begins. If you would prefer to inhibit multiple user mode access until you have had a chance to verify that all went well, see Restricting User Access in Multiple User Mode before entering the FILESERV START command.
- Remove the RESTORE startup parameter from the DMSPARMS file at the earliest opportunity. If you specified the CATBUFFERS startup parameter, you should stop multiple user mode, reset CATBUFFERS to its regular value (or allow it to default), and then resume multiple user mode.