CloudPoint host sizing recommendations
The CloudPoint host configuration depends primarily on the number of workloads and also the type of workloads that you wish to protect. It is also dependent on the maximum number of simultaneous operations running on the CloudPoint server at its peak performance capacity.
Another factor that affects performance is how you use CloudPoint for protecting your assets. If you use the CloudPoint agentless option to discover and protect your assets, then the performance will differ depending on the type of workload.
With agentless, CloudPoint transfers the plug-in data to the application host, performs the discovery and configuration tasks, and then removes the plug-in package from the application host.
Veritas recommends the following configurations for the CloudPoint host:
Table: Typical CloudPoint host configuration based on the number of concurrent tasks
Workload metric | CloudPoint host configuration |
|---|---|
Up to 16 concurrent operational tasks |
CPU: 2 CPUs Memory: 16 GB For example, in the AWS cloud, the CloudPoint host specifications should be an equivalent of a t3.xlarge instance. |
Up to 32 concurrent operational tasks | CPU: 4 - 8 CPUs Memory: 32 GB or more For example, in the AWS cloud, the CloudPoint host specifications should be an equivalent of a t3.2xlarge or a higher type of instance. |
General considerations and guidelines:
Consider the following points while choosing a configuration for the CloudPoint host:
To achieve better performance in a high workload environment, Veritas recommends that you deploy the CloudPoint host in the same location as that of the application hosts.
If you are using the agentless option, Veritas recommends that you allocate enough space to the
/tmpdirectory on the application host. CloudPoint uses this directory for extracting the plug-in configuration files.Depending on the number of workloads, the amount of plug-in data that is transmitted from the CloudPoint host can get really large in size. The network latency also plays a key role in such a case. You might see a difference in the overall performance depending on these factors.
If you wish to configure multiple workloads using the agentless option, then the performance will be dependent on factors such as the network bandwidth and the location of the CloudPoint host with respect to the application workload instances. You can, if desired, bump up the CloudPoint host's CPU, memory, and network configuration to achieve a performance improvement in parallel configurations of agentless application hosts.
In cases where the number of concurrent operations is higher than what the CloudPoint host configuration capacity can handle, CloudPoint automatically puts the operations in a job queue. The queued jobs are picked up only after the running operations are completed.