Home > Storage > PowerScale (Isilon) > Product Documentation > Networking > Dell PowerScale: Network Design Considerations > Load balancing
SmartConnect load balances incoming network connections across SmartConnect zones composed of nodes, network interfaces, and pools. The load-balancing policies are Round Robin, Connection Count, CPU Utilization, and Network Throughput. The most common load-balancing policies are Round Robin and Connection Count, but these policies might not apply to all workloads. It is important to understand whether the front-end connections are being evenly distributed, either in count or by bandwidth. Front-end connection distribution can be monitored with InsightIQ or the WebUI. Because each workload is unique, understand how each load-balancing policy functions, and test the policy in a lab environment before a production roll-out.
The following table lists suggested policies based on the workflow, but these are general suggestions, and might not always be applicable.
Start with Round Robin for a new implementation or if the workload is not clearly defined. As the workload is further defined and based on the Round Robin experience, another policy can be tested in a lab environment.
Load-balancing policy | General or Other | Few clients with extensive usage | Many persistent NFS & SMB connections | Many transitory connections (HTTP, FTP) | NFS automounts or UNC paths |
Round Robin | Yes | Yes | Yes | Yes | Yes |
Connection Count* | Yes | Yes |
| Yes | Yes |
CPU Utilization* |
|
|
|
|
|
Network Throughput* |
|
|
|
|
|
* Metrics are gathered every 5 seconds for CPU Utilization and every 10 seconds for Connection Count and Network Throughput. In cases where many connections are created simultaneously, these metrics might not be accurate, creating an imbalance across nodes.
Note: In contrast to “CPU” and “Throughput” methods where it is not possible to estimate the additional load that a new connection will create with complete accuracy, the “Connection Count” policy is able to accurately account for new connections. Therefore, it does not need to poll to correctly track additional connections. However, it is unaware of client disconnects that occur between poll intervals. Depending on the number and distribution of disconnects, this may cause a temporary connection imbalance across nodes.
As discussed previously, the preceding policies mapping to workloads are general guidelines. Each environment is unique with distinct requirements. Confirm the best load-balancing policy in a lab environment that closely mimics the production environment.