NetBackup™ Web UI VMware Administrator's Guide
- Introducing the NetBackup web user interface
- Monitoring NetBackup
- Managing VMware servers
- Protecting VMs
- Instant access
- Instant rollback
- Continuous data protection
- VM recovery
- Troubleshooting VMware operations
Sizing considerations
This section describes the sizing requirements of the CDP gateway, based on the workload in your environment.
Note:
If the CDP gateway plans to support large number of VMs, it is recommended to deploy the CDP gateway, and the MSDP or media server hosting the storage unit, on different hosts.
Note:
If CDP gateway and MSDP are co-located on the same media server, then CDP service consumes 20% of available memory (RAM) for its internal use. If the CDP gateway is standalone on media server, it consumes 50% of available memory for the same.
You need to size the CDP based on the number of VMs that you want to protect. Consider the requirements described in this section, while calculating requirements for the gateway.
CDP enables you to continuously tap the IOs done by the VMs. NetBackup, by default, uses 10-GB storage space on the staging area per VM. When IO tapping starts, the CDP service starts writing the data into this 10GB storage. Once this storage limit is reached, the CDP service (nbcctd) triggers a backup job to move this data from the gateway to the backup storage.
Out of the total available space on the CDP staging path, by default, NetBackup reserves 25% for usage beyond allocated storage per VM. This storage is common for the subscribed VMs to the gateway. You can reconfigure this value in the nbcct.conf
file.
To configure reserved storage
- Logon to CDP gateway.
- Navigate to the
<NBU install location>/netbackup/nbcct/
directory, and open thenbcct.conf
file in a text editor. - Enter a value against the parameter and CCT_VM_QUOTA_RESERVE_PERCENT
- Restart the nbcctd service.
Storage requirement for the gateway
When NetBackup receives the data from the ESXi IO daemon, it stores the data in the in-memory cache. Recommenced is minimum 160 MB of data for each VM.
For example, you protect 40 VMs in a gateway. So, you need 40*160-MB = 6400-MB RAM. Allocating more RAM increases the in-memory cache size when CDP service starts, ultimately increasing the IO performance of the service.
Similarly, to stage 40 * 10-GB = 400-GB (75%) + 134GB (25%) reserved, that is approximately 540 GB space you need to have on the staging area.
Increasing per VM storage allows to NetBackup to backup more data per backup job. Increasing reserved storage for the CDP gateway lets you receive more data without any interruption to the protection. Note that even when the staging path is fully occupied, it does not affect the applications inside the VM. NetBackup catches up the data produced by applications during that time, moves it to the backup storage in the subsequent backup job(s).
Note:
If NFS is used for the staging area, minimum required throughput is 100 MB/sec.
When you start using the CDP feature, it is important to observe the system and tune according to your business demand, add hardware configuration to maximize the protection and performance. First, you can use default values and start subscribing the VMs according to the requirements mentioned in this section. You should check the following:
Number of immediate backup jobs that the CDP service triggers due to staging storage full condition.
You can check the CDP backup engine notifications on NetBackup web UI.
Underlying provisioned storage performance. Like the NetBackup installation disk, CDP staging area, and MSDP storage disk(s).
Network utilization and available bandwidth.
CPU and memory consumption when receiving data from the ESXi, and when the backup jobs are running.
Note:
If you observe slow IOs from the IO daemon, check network bandwidth and system RAM. You can increase the in-memory cache size using CCT_POOL_SIZE_QUOTA_PERCENTAGE parameter in the nbcct.conf
file.