
RecoverPoint for Virtual Machines 5.3.3.2 Scale and Performance Guide
Considerations for choosing the vRPA configuration
The following are considerations for choosing the vRPA configuration:
- It is recommended to use only the predefined configurations.
- The 4 CPUs, 16 GB RAM ("Silver") configuration is appropriate for most situations. If low throughput is expected, use a 2 CPUs, 16 GB RAM ("Bronze") configuration. For synchronous replication, or when deduplication is enabled, or where reducing latency is critical, use the 8 CPUs, 16 GB RAM ("Gold") configuration.
- It is possible to manually allocate more vCPU and memory to deployed vRPAs.
- Assigning more memory/vCPU may increase the performance of the vRPA.
- For some workloads, weak vRPAs are more cost-effective than strong ones. For example, as shown in Tables 2 and 3, the IOPS and throughput capabilities of an 8 vCPUs, 16 GB RAM vRPA are less than twice the capabilities of a 4 vCPUs, 16 GB RAM vRPA.
- The more memory the vRPA has, the stronger and longer peaks of write I/Os it can handle before entering high load.
- In a vRPA cluster, the first two vRPAs are candidates to run cluster control. The cluster control vRPA runs more logic and therefore requires more CPU and memory resources, especially in scaled environments with many VMs. Hence, in scaled environments, it is advisable to deploy strong vRPAs at least for the first two vRPAs in the cluster.
- It is recommended to upgrade vRPA1 and vRPA2 in all vRPA clusters to 32 GB RAM to support large-scale, multicluster environments, with over 2000 VMs in inventory,
Table 1 summarizes the main factors to consider when selecting a vRPA configuration.
| vRPA configuration | Supported replication use case |
|---|---|
| 2 CPUs, 16 GB RAM
("Bronze") |
|
| 4 CPUs, 16 GB RAM
("Silver") |
|
| 8 CPUs, 16 GB RAM
("Gold") |
|
Example: Choosing vRPA configuration
I must replicate a few VMs, but I do not know the workload that they generate. Which vRPAs configuration should I choose when I deploy the vRPA cluster?
A possible strategy is to deploy a small vRPA cluster of two 2 vCPUs, 16 GB RAM vRPAs because these vRPAs are the most cost-effective for throughput. After protecting the VMs, check the throughput statistics of the CGs and examine high loads. Depending on the loads, adjust the vRPA configuration appropriately, and even add vRPAs to the vRPA cluster, if required.
Please provide ratings (1-5 stars).
Please provide ratings (1-5 stars).
Please provide ratings (1-5 stars).
Please select whether the article was helpful or not.
Comments cannot contain these special characters: <>()\