Restore process
Understanding how the restore process works is a helpful first step in deciding which logs to gather for a particular issue. The restore process differs depending on whether you restore an image from tape or from disk.
Figure: Restore from tape process flow illustrates a restore from tape.
- The (1) NetBackup request daemon (bprd) receives a restore request. This request can be initiated from the Backup, Archive, and Restore user interface or from the (2) command line (bprestore).
- The bprd process launches two child processes: MAIN bprd and MPX-MAIN-bprd. The MAIN bprd process is used to identify images and media, while the MPX-MAIN-bprd process manages the restore operation. For simplicity's sake, these three processes are all referred to here as bprd.
- The bprd service communicates with the (3) NetBackup Database Manager program (bpdbm) to get the information that is required to restore the files that have been requested.
- Once it has the information it needs, bprd communicates with (4) bpjobd, and the job is added to the job list in the jobs database. The job is now visible in the Activity Monitor. It may show as "Active" even before resources are acquired.
- The bprd service goes through Private Branch Exchange (PBX) and the NetBackup Legacy Network (vnetd) to start the (5) NetBackup backup and restore manager (bpbrm).
- The bpbrm service starts the (6) tape management process (bptm) and provides the media information that is required for the restore. It also starts the (7) Tape Archive program (tar) on the client (through PBX and vnetd) and creates a connection between tar and bptm.
- The bptm process sends a resource request to the (8) NetBackup Job Manager (nbjm) through PBX and vnetd.
- The nbjm process sends the resource request to the (9) NetBackup Resource Broker (nbrb), which queries the (10) Enterprise Media Manager (nbemm). Once the resources have been allocated, nbrb notifies nbjm, which notifies bptm.
- The bptm process makes a mount request to the (11) logical tape interface daemon (ltid). The ltid service calls on the (12) robotic drive daemon (txxd, where xx varies based on the type of robot being used). The txxd daemon communicates the mount request to the (13) robotic control daemon (txxcd), which mounts the media.
- The bptm process reads the data to be restored from the media and delivers it to tar.
- The tar process writes the data to the client disk.
- When the restore is completed, bptm unmounts the media and notifies nbjm. The job now appears as "Done" in the Activity Monitor.
Some additional logs that are not included in the restore process flows but that can be of use in resolving restore problems include: reqlib, daemon, robots, and acsssi.
Figure: Restore from disk process flow illustrates a restore from disk.
- The (1) NetBackup request daemon (bprd) receives a restore request. This request can be initiated from the Backup, Archive, and Restore user interface or from the (2) command line (bprestore).
- The bprd process contacts the (3) NetBackup Database Manager program (bpdbm) to identify the files, the client, and the media information for the restore.
- The bprd process initiates a (4) child bprd process. The child bprd process makes a call to the (5) Enterprise Media Manager (nbemm) to verify that the disk storage unit is available.
- The child bprd process communicates with (6) bpjobd to allocate a jobid. The restore job is now visible in the Activity Monitor.
- The bprd process starts the (7) NetBackup backup and restore manager (bpbrm) on the media server, through Private Branch Exchange (PBX) and the NetBackup Legacy Network Service (vnetd).
- The bpbrm service uses PBX and vnetd to establish a connection with the (8) Tape Archive program (tar) on the client system. It also starts the (9) tape management process (bptm).
- The bptm process makes a call to bpdbm (through PBX and vnetd) to get the fragment information and then mounts the disk.
- The bptm process reads the backup image from the disk and streams the requested data to tar.
- The tar process commits the data to the storage destination.
Each of the processes that is involved in a restore has an accompanying log file. These logs can be consulted to diagnose any issues that you encounter with your restore.