Veritas CloudPoint Administrator's Guide
- Getting started with CloudPoint
- Section I. Installing and configuring CloudPoint
- Preparing for installation
- Deploying CloudPoint
- Deploying CloudPoint in the AWS cloud
- Using plug-ins to discover assets
- Configuring off-host plug-ins
- AWS plug-in configuration notes
- Google Cloud Platform plug-in configuration notes
- Microsoft Azure plug-in configuration notes
- HPE RMC plug-in configuration notes
- NetApp plug-in configuration notes
- Hitachi plug-in configuration notes
- InfiniBox plug-in configuration notes
- About CloudPoint plug-ins and assets discovery
- Configuring the on-host agents and plug-ins
- Oracle plug-in configuration notes
- Protecting assets with CloudPoint's agentless feature
- Preparing for installation
- Section II. Configuring users
- Section III. Protecting and managing data
- User interface basics
- Indexing and classifying your assets
- Protecting your assets with policies
- Tag-based asset protection
- Replicating snapshots for added protection
- Managing your assets
- About snapshot restore
- Single file restore requirements and limitations
- Additional steps required after a SQL Server snapshot restore
- Monitoring activities with notifications and the job log
- Protection and disaster recovery
- Section IV. Maintaining CloudPoint
- CloudPoint logging
- Troubleshooting CloudPoint
- Working with your CloudPoint license
- Managing CloudPoint agents and plug-ins
- Upgrading CloudPoint
- Uninstalling CloudPoint
- Section V. Reference
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 plugin data to the application host, performs the discovery and configuration tasks, and then removes the plugin package from the application host. Therefore, database applications such as Oracle and Microsoft SQL Server will require a higher capacity configuration, as compared to other assets.
Veritas recommends the following configurations for the CloudPoint host:
Table: Typical CloudPoint host configuration
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 t2.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 t2.2xlarge 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
/tmp
directory on the application host. CloudPoint uses this directory for extracting the plugin configuration files.Depending on the number of workloads, the amount of plugin 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.