Balance load on Additional polling engines in an HA pool

Engine Load Balancing (ELB) distributes workload across engines in a high‑availability (HA) pool to improve performance, stability, and resource utilization. ELB supports automatic and manual node rebalancing, health‑based decision making, and configurable weighting for key performance metrics. Use ELB to maintain consistent engine health and prevent uneven load across your environment.

Requirements

To use Engine Load Balancing (ELB), ensure the following requirements are met:

• Additional polling engines must be part of the same HA pool.

• Additional polling engines must be monitored as nodes.

• ELB must be enabled at the global level.

• ELB must be enabled for each HA pool where load balancing is required.

Limitations

• Standalone Additional polling engines are not supported. ELB is available only for engines that belong to an HA pool and are monitored as nodes.

• Re-assigning Agent‑polled nodes are not supported. If a pool includes Agent-polled nodes, these nodes remain on the original polling engine.

• Remote Collector–polled nodes are not supported.

• Advanced dependency‑aware balancing is not available. ELB does not evaluate complex dependencies.

• Database latency and additional performance metrics are not included. Only the current Engine Health Score metrics (Polling Completion, CPU Load, Memory, and Loss) are used for health‑based decisions.

Enable the global engine load balancing

1. Log in to the SolarWinds Platform Web Console as an administrator and go to Advanced Configuration. Adjust the SolarWinds Platform Web Console URL as follows: [hostname]/Orion/Admin/advancedconfiguration/global.aspx

2. Locate EnableEngineLoadBalancing.

3. Select the EnableEngineLoadBalancing box.

4. Save your changes.

After you enable ELB, engine load balancing settings and related feature controls become available. Load balancing begins based on the settings you configure. See Enable engine load balancing for individual pooland Configure load balancing settings.

When ELB is enabled, a background service runs at regular, configurable intervals. It reviews each ELB‑enabled HA pool, calculates the Health Score for every engine, and automatically redistributes nodes to healthier engines when rebalancing is required.

Enable engine load balancing for individual pool

After you enable ELB globally, enable it for individual HA pools.

1. In the SolarWinds Platform Web Console, click Settings > My Deployment.

2. Go to the Pools tab.

3. Select a pool and click Enable Engine Load Balancing.

If the pool includes multiple engines, their load is automatically balanced, according to ELB settings. See Configure load balancing settings.

Run load balancing manually for a pool

After you enable engine load balancing both globally and for a pool, you can run load balancing manually.

1. In the SolarWinds Platform Web Console, click Settings > My Deployment.

2. Go the Pools tab.

3. Select a pool and click Run Engine Load Balancing.

Nodes monitored by each polling engines are redistributed based on the engine health score.

Monitor ELB

1. In the SolarWinds Platform Web Console, click Alerts & Activity > Message Center.

2. Ensure Show Audit Events is selected.

3. To review ELB activity, filter for Engine Load Balancing Execution under Filter Audits: Action type, and select Apply.

   Audit messages include:

   1. ELB start time

   2. User who initiated ELB

   3. Engine Health Score

   4. Number of nodes re-assigned

4. To see details on which nodes were re-assigned, filter for Engine Load Balancing Nodes Reassigned.

Configure load balancing settings

You can configure how often ELB runs, how engine health is calculated, and how many nodes can be reassigned in a single ELB run.

1. In the SolarWinds Platform Web Console, click Settings > My Deployment.

2. Go the Pools tab.

3. Edit the load balancing Settings:

   • Task interval: Set the Engine Load Balancing Task Interval to control how frequently ELB runs.

   • Metric weighing: Adjust the weights for CPU load, percent memory used, polling completion, and percent loss.

     • Polling engines must be monitored as nodes, otherwise only polling completion is available.

     • If some metrics are unavailable, ELB uses the available metrics and adjusts the weights.

     • If no metrics are available, the engine's health score is set to null and the engine is skipped.

   • Health score thresholds: Set the health score threshold and health score safe margin.

     • Engines below the threshold are considered unhealthy. As a result, some nodes may be reassigned away.

     • Engines must exceed the health score threshold + the safe margin to qualify for receiving additional nodes.

   • Node reassignment percentage: Set the percentage of nodes that can be re-assigned during each run.

4. Save your changes.