Veritas NetBackup™ Logging Reference Guide
- Using logs
- About unified logging
- About legacy logging
- About global logging levels
- Troubleshooting error messages in the NetBackup Administration Console
- Backup process and logging
- Media and device processes and logging
- Restore process and logging
- Advanced Backup and Restore Features
- Storage logging
- NetBackup Deduplication logging
- OpenStorage Technology (OST) logging
- Storage lifecycle policy (SLP) and Auto Image Replication (A.I.R.) logging
- Snapshot technologies
- Locating logs
- Java-based administration console logging
VMware backup
The following shows a VMware backup process.
The basic processing steps for a VMware backup operation are the following:
- The Policy Execution Manager (nbpem) triggers a backup job when the policy, schedule, and virtual machine are due and the backup window is open. The nbpem process, the Job Manager (nbjm), the Resource Broker (nbrb), and the Enterprise Media Manager (nbemm) together identify the resources (media server, storage unit, etc.) for the backup operation.
- For a VMware Intelligent Policy (VIP), you can throttle the VMware resources that are used in the vSphere environment. For example, you can limit the resources to four concurrent backup jobs running from a vSphere datastore. This level of control tunes the number of backups to minimally influence the user and application experience on the vSphere platform.
- nbpem uses nbjm to contact the selected media server and to start the Backup and Restore Manager (bpbrm) on it. A snapshot job (also referred to as the parent job) goes active in the Activity Monitor.
- nbjm starts an instance of bpbrm through the client service (bpcd) on the media server. bpbrm starts the Frozen Image Snapshot (bpfis) through the client service (bpcd) on the VMware backup host. bpfis creates a snapshot of the VM data by using vCenter or ESX host depending on the configured credential servers.
bpfis armed with vADP contacts the vSphere host (vCenter) or the ESX/ESXi host for which credentials are stored in the NetBackup database and initiates the snapshot for the VM. For multiple VMs, bpbrm starts bpfis for each VM so that the snapshot operations occur in parallel. As in step 2, you can control the number of concurrent snapshots for a VIP by setting VMware resource limits in NetBackup. bpfis contacts the vSphere host by using the standard SSL port (the default is 443).
- bpfis contacts the Request Manager (bprd) to request transfer of bpfis state files from the VMware Backup Host to the master server.
- bprd requests bpcd on the VMware Backup Host to send a list of bpfis state files. bprd copies each state file from the VMware Backup Host to the master server.
- bpfis sends snapshot information and completion status to bpbrm. bpbrm reports the snapshot information and status to nbjm. nbjm relays the information and status to nbpem.
- nbpem submits a child job for the backup to nbjm, with a file list derived from the snapshot information. nbjm starts bpbrm to back up the snapshot.
- bpbrm uses bpcd to start bpbkar on the VMware Backup Host.
- The backup and archive manager (bpbkar) loads the Veritas Mapping Services (VxMS) which loads the VMware Disk Development Kit (VDDK) APIs. The APIs are used for reading from the vSphere datastore. VxMS maps the stream during run-time and identifies the contents of the vmdk file. bpbkar uses VxMS to send the file catalog information to bpbrm, which relays it to the database manager bpdbm on the master server.
- bpbrm also starts the process bptm (parent) on the media server.
The following shows the operation of the Veritas V-Ray within VxMS:
Veritas V-Ray within VxMS generates the catalog of all the files inside the VMDK from both Windows and Linux VMs. The operation occurs while backup data is being streamed. bpbrm on the media server sends this catalog information to the master server.
The file system inode level also identifies unused and deleted blocks. For example, if the application on VM allocates 1 TB of space for a file, of which only 100 GB is currently used, the backup stream includes only that 100 GB. Similarly, if you delete a 1 TB file that was fully allocated in the past, VxMS skips the deleted blocks (unless the blocks are now allocated for a new file) from the backup stream. This optimization not only speeds up the backup stream, but reduces needed storage even when deduplication is not enabled.
If the source side deduplication feature is enabled, the VMware backup host does the deduplication. The NetBackup deduplication plug-in using the mapping information that VxMS generates and sees the actual files in the file system within the VMDK. This V-Ray vision is established by the NetBackup deduplication plug-in that loads a dedicated stream handler that understands the VxMS mapping info.
Because these operations occur on the VMware backup host, the ESX resources and the VM resources are not used. This setup is true off-host backup with no burden on the production vSphere. Even the source side deduplication occurs in an off-host system.
- If the media server is the VMware Backup Host, bpbkar stores the snapshot-based image block-by-block in shared memory on the media server. If the media server is backing up a separate VMware Backup Host that is not the media server, the bptm process on the server creates a child process of itself. The child uses socket communications to receive the snapshot-based image from the VMware Backup Host and stores the image block-by-block in shared memory.
- The original tape manager (bptm) process takes the backup image from shared memory and sends it to the storage device (disk or tape).
- bptm sends backup completion status to bpbrm, which passes it to nbjm and nbpem.
- nbpem tells nbjm to delete the snapshot. nbjm starts a new instance of bpbrm on the media server, and bpbrm starts a new instance of bpfis on the VMware Backup Host. bpfis deletes the snapshot on the vSphere environment. bpfis and bpbrm report their status and exit.