Setting global limits on the use of RHV resources
You can control the number of simultaneous backups that can be performed on an RHV resource type. The settings apply to all NetBackup policies for the currently selected primary server.
For example, to avoid overloading the overall RHV cluster, you can place a limit on the number of concurrent backup jobs per RHV Cluster. To control input output overhead on the storage domain array, you can limit the number of concurrent backups per storage domain.
Resource limits available for RHV:
Backup Jobs per Cluster
Backup Jobs per DataCenter
Backup Jobs per StorageDomain
To set limits on the use of RHV resources
- On the left, click Workloads > RHV.
- On the top right, click RHV settings > Resource limits.
For each resource type, the default is 0, (No Limit).
- Select the resource type you want to change and then click Edit.
- Choose from the following options.
Set a global limit for a VMware resource type.
Locate the Global setting and select the Limits value that you want to apply.
This value limits the number of simultaneous backups that are performed for the resource type.
Set a limit for a specific VMware resource.
Click Add.
Provide a name for the resource limit.
Select the Limits value that you want to apply.
This value limits the number of simultaneous backups that are performed for the specific resource.
The following example illustrates how these limits control simultaneous backups. The settings must be done according to RHV configuration in your environment.
When NetBackup connects to the RHV environment for backup, it makes 1 connection per disk present on the VM. So, if a VM has 2 disks then NetBackup would make 2 connections with the RHV node.
So, consider a case where RHV manager is managing 2 clusters with 2 nodes in each cluster. Let's consider every node is hosting 20 VMs with 2 disks on each VM.
When the job runs, there would be 80 concurrent jobs start if no Resource Limit is set which is default behavior. RHV recommends up to 10 disk connections concurrently per node in the cluster. In the example of each VM with two disks, 5 VMs per node can be ideally backed up concurrently. Hence, in this case of a cluster with two nodes, it is recommended to backup upto 10 VMs concurrently. When the Backup Jobs per Cluster is set to 10, this limit is enforced.
Since one data center can manage multiple clusters, Backup Jobs per DataCenter resource limit could be higher than Backup Jobs per Cluster. You need to determine the values by evaluating the effect of backup on overall environment.
Storage domain in RHV can server multiple clusters in a data center. It serves to protection VM as well as backup. Its performance is dependent on type of storage technology - FC, iSCSI, NFS, gluster etc. Hence limit on storage domain using Backup Jobs per StorageDomain can be set based on characteristics of storage domain technology and limit can be higher than Backup Jobs per Cluster.